 

 

bug A

1

. .
w .

4 17,: Q I}. ..
xii? 0‘1
ant»... guard

1 7|:vns‘n‘rl 3’]
1.. a

.1. A

1.4.1.5.‘35

. . It. :3. v

I. 7131‘ .5...
dhrunif‘.
: .2141.

.3, 11‘! .

 

LIBRARY
Michigan State
University

This is to certify that the
dissertation entitled

START UP OF ANAEROBIC BIOREACTOR LANDFILLS
IN COLD CLIMATES BY INTERMITTENT AIR
INJECTION

presented by

Reem Raji Musleh

has been accepted towards fulﬁllment
of the requirements for the

Ph.D degree in Environmental Engineering

 

 

Major Professor’s Signature

 

“[3012009?

 

Date

MSU is an Afﬁnnative Action/Equal Opportunity Institution

~—-o---o-o-n--n--o-o-a----«-n-

- _«-.-._._.—.--.--.-u--n--o-n-.-a-.-n---.au_.g.—u-n-a-n-n-o--—.—.:—

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 p:/ClRC/DateDue.indd-p.1

 

START UP OF ANAEROBIC BIOREACTOR LANDFILLS IN COLD
CLIMATES BY INTERMITTENT AIR INJECTION

By

Reem Raji Musleh

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Civil and Environmental Engineering

2005

ABSTRACT

START UP OF ANAEROBIC BIOREACTOR LANDFILLS IN COLD CLIMATES BY
INTERMITTENT AIR INJECTION

By

Reem Raji Musleh

Intermittent air injection may be one of the most economical solution to enhance waste
temperature without toxicity to methanogenesis. Waste placed in landﬁll at lower
temperatures may remain at lower temperature, leading to slow startup of anaerobic
digestion process in biodegradation in landﬁlls. In this study, waste placed at ambient
temperature of 3.2i0.3 °C (range between -30°C to +30°C), was subjected to leachate
injection to increase waste moisture content. This has resulted in waste temperature of
10.4i1 .3°C. After leachate injection, the was subjected to intermittent horizontal air
injection increasing the temperature by 10.2:2.6°C over the entire cell using 2.5hp
compressor.

The objectives of this study were to quantify the relationship between amount of heat
consumed and amount of heat production (known as heat generation factor), to determine
Point oxygen consumption rates (i.e., that is oxygen consumption rate at particular
location in the landﬁll not normalized to the waste mass), and to evaluate the impact of
air injection on methanogenesis process.

The average heat generation factor under ﬁeld conditions was 341:84 kJ/mole, using
dynamic calorimetric method. First order POCR averaged 00867100422 h-l after 15

days of air injection and increased to O.2450i0.0843 h-l after 21 days of air injection.

This study demonstrated that air injection is key to a successful startup of anaerobic
digestion of bioreactor landﬁlls in colder climates. The average rate of increase in
percent methane at all locations exposed to air in the landﬁll increased signiﬁcantly from
0.15:0.042 °C day-1 to O.68:t0.15°C day-1. This was accompanied by a decrease in the
median of hydrogen concentration from 4,085 ppm to 953 ppm. The lag time also

decreased from 137 days to 15 days, in response to air injection.

Cepyright by
REEM RAJI MUSLEH

2005

DEDICATION

To my family

ACKNOWLEDGEMENTS

I want to convey my thanks to my advisor Dr. Syed Hashsham and to Dr. Xianda Zhao
for the guidance and support they provided through this research. I also convey my
gratitude to my committee members Dr. Thomas C. Voice, Dr. Susan Masten, and Dr.

Stephen Boyd.

This project was supported by grants from the Environmental Research and Education
Foundation with funding from the Department of Energy (under special project number
DE_FG36-016011096) and from Waste Management, Inc. I am especially thanka to
Dr. Ed Repa, whose initiative and continuous support was critical for this project. I
would like to extend my gratitude for the staff at Northern Oaks Waste Recycling Facility
at Harrison, M1 for their cooperation. I would also like to acknowledge the support

provided by the Fullbright Fellowship in support of my graduate studies.

It is also fair to mention the name of my colleague Irfan Aslam, with whom I could

discuss my research problems and get a meaningful critique and ideas to work on.

I feel obliged to mention here my uncle Tank and his family as I was privileged to enjoy
their love, company and continuous support since my ﬁrst day at MSU. I want to thank
my parents, and my sister for their continuous motivation and support that helped me
throughout my studies. Special thanks to all my friends at MSU who were always there

for me: Rana, J ameel, Irfan and Dominique.

vi

 

TABLE OF CONTENTS

 

 

 

 

 

List of Tables - ix
List of Figures- - xii
1. INTRODUCTION 1
1.1. Temperature .................................................................................................. 2
1.2. Heat Generation Factor ................................................................................. 3
1.3. Oxygen Consumption Rate ........................................................................... 3
1.4. Impact of Air Injection on Methanogenesis .................................................. 4
1.5. Heterogeneity ................................................................................................ 5
1.6. Objectives ..................................................................................................... 8
2. LITERATURE REVIEW - 10
2.1 . Temperature .................................................................................................. 10
2.2. Enhanced air diffusion .................................................................................. 10
2.2.1 Air injection .......................................................................................... 12
2.2.2. Radius of inﬂuence .............................................................................. 14
2.3. Heat Generation Factor ................................................................................. 14
2.4. Oxygen Consumption Rate ........................................................................... 17
2.4.1 Effect of process variables on aerobic degradation .............................. 18
2.5. Anaerobic Digestion of Solid Waste ............................................................. 20
2.5.1 Effect of temperature on anaerobic digestion ....................................... 22
2.5.2. Inhibition of methanogenesis due to low pH and high VFA ............... 23
2.5.3. Inhibition of methanogenesis due to oxygen ....................................... 24
2.5.4. Hydrogen and anaerobic digestion process .......................................... 24
3. SITE DESCRIPTION AND METHODS 27
3.1 Site Description and Bioreactor Landﬁll Cell Construction .......................... 27
3.2. Monitoring Devices ...................................................................................... 28
3.2.1. Automated and onsite measurements ................................................... 33
3.2.2. Gas and leachate sampling protocol for off-line analysis .................... 34
3.3. Analytical Methods ....................................................................................... 34
3.3.1. Hydrogen .............................................................................................. 35
3.3.2. Oxygen, nitrogen, carbon dioxide and methane .................................. 35
3.3.3. Electrical conductivity, pH, COD, VOC ............................................. 35
3.4. Experimental Procedures .............................................................................. 36
3.4.1.. Performance of intermittent air injection system (Objective 1) ........... 38
3.4.2. Oxygen consumption rate (Objective 2) .............................................. 41
3.4.3. Evaluation of anaerobic digestion process after startup or air perturbation
(Objective 3) .................................................................................................. 41

vii

3.5. Statistical Methods ........................................................................................ 42

3.5.1. Normality assumption .......................................................................... 43
3.5.2. Test difference between two independent samples .............................. 45
3.5.3. Test difference between two dependent samples ................................. 47
3.5.4. Tests the linearity of a relationship between two variables ................. 49
3.6. Method to determine heat generation factor ................................................. 51
3.7. Method to determine oxygen consumption rate ............................................ 57
3.7.1. Oxygen concentration vs. time curve ................................................... 58
3.7.2. Calculation of liquid phase oxygen concentration ............................... 59

3.7.3. Calculation of kinetic coefﬁcients for POCR using Monod equation .62

 

 

 

4. RESULTS AND DISCUSSION - - 70
4.1. Evaluation of Intermittent Air Injection in Northern Oaks Bioreactor Landﬁll
.............................................................................................................................. 7 1
4.1.1. Establishment of base line temperature before air injection ................ 71
4.1.2. Effect of air injection on waste temperature ........................................ 79
4.1.3. Effect of moisture on rate of temperature increase .............................. 90
4.1.4. Temperature response after air injection tests ..................................... 93
4.1.5. Heat Generation Factor ........................................................................ 97
4.1.6. Change in oxygen, nitrogen, carbon dioxide and methane concentration
due to air injection ......................................................................................... 109
4.1.7. Impact of air injection on leachate characteristics .............................. 111
4.2. Oxygen Consumption Rate ........................................................................... 119
4.2.1. Operational conditions ......................................................................... 120
4.2.2. Fitting POCR data to Monod equation ................................................ 122
4.2.3. Fitting ﬁrst order kinetics to POCR data ............................................. 126
4.2.4. Factors affecting POCR ....................................................................... 129
4.2.5. Comparison between POCR and OCR from the landﬁll ..................... 133
4.3. Evaluation of methanogensis ........................................................................ 136
4.3.1. Methane generation and hydrogen accumulation during startup ......... 136
4.3.2. Methane generation and hydrogen accumulation after stopping air
injection ............................................................................................................. 149
4.3.3. Factors affecting methanogensis establishment during bioreactor landﬁll
startup ............................................................................................................. 159
4.3.4. Factors affecting methanogenesis establishment aﬁer air injection was
stopped ........................................................................................................... 164
4.3.5. Comparison of methanogensis establishment between startup and after air
injection .......................................................................................................... 167
5. CONCLUSIONS - - - -- -- 169
6. ENGINEERING SIGNIFICANCE - 172
7. REFERENCES - -- 175

 

viii

LIST OF TABLES

Table 1-1: Temperature measured in the landﬁll on 5/9/03 ..................................... 7
Table 2-1: Air ﬂow rate and temperature change as reported in aerobic bioreactor
landﬁlls ...................................................................................................................... 13
Table 2-2: OCR reported for solid waste ................................................................... 20

Table 3-1: Chronological order of ﬁlling the Northern Oaks Bioreactor Landﬁll ....29
Table 3-2: Air injection test 1 (lower lift) — Percent ﬂow in each injection line ....... 40

Table 3-3: Air injection test 2 (upper 1iﬁs)—injected air volume distribution among the
injection lines ............................................................................................................. 40

Table 3-4: Comparison between parametric and nonparametric tests ....................... 43

Table 3-5: Calculation of the standard error of the difference and the degree of freedom
for the t-test under homogeneous and non-homogeneous variances ......................... 46

Table 4-1: Kruskal-Wallis Test for testing differences between lifts waste lifts with
respect to a) ambient temperatures to which waste was exposed and b)waste temperature

.................................................................................................................................... 73
Table 4-2: AN OVA test for waste temperature differences for lifts 1 through 4 after

leachate recirculation (9/5/03) ................................................................................... 77
Table 4-3: Operational conditions for each air injection event .................................. 80
Table 4-4: Rate of temperature change for air injection events in each lift ............... 81

Table 4-5 Air injection events for which the heat balance was conducted and their
operational parameters ............................................................................................... 98

Table 4-6: Calculation results of input parameters that varied for each air injection

.................................................................................................................................... 99

Table 4-7: Input parameters for the calculation of heat balance equation ................. 100
Table 4-8: Heat loss, production and accumulation in the bioreactor landﬁll ........... 105
Table 4-9: Average heat generation factor ............................................................... 106

ix

Table 4-10: Comparison between laboratory values for heat generation factor in

laboratory (single substrates) and ﬁeld scale (solid waste) ....................................... 106
Table 4-11: Total heat loss and gain and the percent contribution of each term of the heat
balance equation ......................................................................................................... 108
Table 4—12: Average concentrations of VOC in lift 1 ................................................ 113
Table 4-13: Henry’s constant for VOCs found in leachate samples from Northern Oaks
bioreactor landﬁll ...................................................................................................... 1 18
Table 4-14 Ratio of RL/RT VOCs found in leachate samples from Northern Oaks
bioreactor landﬁll for kg/kl=100 .............................................................................. 119
Table 4-15: Sampling locations, temperature, moisture content and oxygen concentration
before stopping air injection at which POCR was conducted (Test 1) ...................... 121
Table 4-16: Sampling locations, temperature, moisture content and oxygen concentration
before stopping air injection at which POCR was conducted (Test 2) ...................... 122
Table 4-17: Comparison between the range of data used and the required to estimate
Monod kinetics parameters ........................................................................................ 124
Table 4-18: First Order POCR ................................................................................... 127

Table 4-19: Tests of Normality for POCR, temperature, moisture content and background
oxygen concentration ................................................................................................. 128

Table 4-20: Wilcoxon Signed Ranks Test for comparison between POCR tests 1 and 2
.................................................................................................................................... 129

Table 4-21: Paired samples t-test for comparison between tests 1 and 2
.................................................................................................................................... 131

Table 4-22: pH, COD, and temperature at sampling locations during landﬁll startup
.................................................................................................................................... 139

Table 4-23: Hydrogen during and after lag time, initial and ﬁnal methane concentration,
rate of increase in percent methane and lag time at sampling locations during landﬁll
startup ......................................................................................................................... 141

Table 4-24: ANN OVA test to determine differences between the different lifts in rate of
increase in percent methane
.................................................................................................................................... 143

Table 4-25 Normality tests for methane concentration, hydrogen concentration and rate of
increase in percent methane
.................................................................................................................................... 144
Table 4-26: t-test to test the difference between upper and lower lifts with respect to the
rate of increase in methane percent
.................................................................................................................................... 144

Table 4—27: Wilcoxon Signed Ranks Test to study the difference in hydrogen
concentration between the two periods: before and lag time
.................................................................................................................................... 146

Table 4-28: Kruskal Wallis Test to study the difference in hydrogen concentrations
between the different lifts
.................................................................................................................................... 147

Table 4-29: Wilcoxon Signed Ranks Test to test differences between hydrogen

concentration during and after lag time for each lift
.................................................................................................................................... 148

Table 4-30: Waste temperature, hydrogen concentration, lag time, ﬁnal methane
concentration and rate of increase in percent methane for locations exposed to air
.................................................................................................................................... 151

Table 4-31: Normality tests for concentration of methane, ﬁnal methane concentration,

hydrogen concentration and rate of increase in percent methane in the landﬁll as a whole
.................................................................................................................................... 152

Table 4-32: Tests of normality per lift for rate of increase in percent methane, methane
concentration and hydrogen concentration ................................................................ 153

Table 4-33: ANN OVA table to test the difference in ﬁnal methane concentration in

between lifts 1, 2 and 3 and in rate of increase in percent methane for lifts 1, 2, 3 and 4
.................................................................................................................................... 154

Table 4-34: Test statistics for Wilcoxon Signed Ranks Test
.................................................................................................................................... 156

Table 4-35: Kruskal Wallis Test to test difference between hydrogen concentrations
among the different lifts ............................................................................................ 157

Table 4-36: Wilcoxon Signed Ranks Test to study the difference in hydrogen
concentration between before and after lag time ....................................................... 157

xi

LIST OF FIGURES
Figure 1-1: Ambient Temperature during waste placement ......................................... 7

Figure 3-1: Bioreactor landﬁll cell at Northern Oaks Recycle and Disposal Facility. A.

The layout of the bioreactor landﬁll cell, B. View of the cell from the north during ﬁlling,

C. View of the cell from north after ﬁlling was completed. ....................................... 28
Figure 3-2 Layout of bioreactor landﬁll at NORDF .................................................... 30
Figure 3-3: Gas sampling port ..................................................................................... 31
Figure 3-4: Leachate sampling port ............................................................................. 32
Figure 3-5: Automata TDR moisture probe and hand held reader .............................. 32
Figure 3-6: Gro.point TDT moisture probe ............................................................... 33

Figure 3-7: Relative location of the weather station on site (This picture was taken from

east side of the bioreactor landﬁll) ............................................................................... 33
Figure 3-8 Overview of experiments timeline and sampling events ........................... 37
Figure 3-9: Gas concentration change after stopping air injection location G143 ...... 58

Figure 3-10: A hypothetical example for determination of the initial rate for Lineweaver-

Burk plot ...................................................................................................................... 68
Figure 4-1: Average waste temperature vs. Average ambient temperature waste is

exposed to in each lift .............................................................................................. . ....72
Figure 4-2: Initial waste temperature in the landﬁll .................................................... 74
Figure 4-3: Gas composition alter waste coverage ...................................................... 76

Figure 44 Average waste temperature in each lift and air temperature during leachate
recirculation ................................................................................................................. 78

Figure 4-5: Average waste temperature in ﬁve lifts in Northern Oaks bioreactor landﬁll
...................................................................................................................................... 82

xii

Figure 4-6: Temperature increase in waste during air injection in L14 for lift2 ........ 83
Figure 4-7: Temperature change during air injection test 1 ......................................... 84
Figure 4-8: Rate of waste temperature increase during test 1 ...................................... 85
Figure 4-9: Temperature change during air injection test 2 ......................................... 87

Figure 4-10: Average rate of waste temperature increase during test 2 while injecting air

through liftl ................................................................................................................ 88
Figure 4-11: Average rate of waste temperature increase during test 2 while injecting

through lifts 3 and 4 .................................................................................................... 89
Figure 4-12: Comparison of the air injection volume between events 1 and 2 .......... 91

Figure 4—13: Comparison of rates of waste temperature increase between air injection
testl and test 2 when injecting air through lift 4 .......................................................... 91

Figure 4-14: Comparison of the rate of waste temperature increase between test 1 and test
2 when injecting air through lift 1 ................................................................................ 92

Figure 4-15: Waste and air temperature after stopping second air injection event ..... 93

Figure 4-16: Average temperature in the landﬁll after stopping air injection tests 1 and 2
...................................................................................................................................... 94

Figure 4-17: Average temperature in lifts 4 and 5 after stopping air injection test 1 with
leachate recirculation periods marked ......................................................................... 95

Figure 4-18: Average temperature in lift 3 after stopping air injection test 1 with leachate
recirculation period through lift 4 marked ................................................................... 96

Figure 4-19: Determination of the rate of temperature increase in the bioreactor landﬁll
for air injection event 9 .............................................................................................. 101

Figure 4-20: Change in gas phase concentration during air injection at location G211 ..
.................................................................................................................................... 110

Figure 4-21 Change in gas phase concentration during the initial hours of air injection at
location G21 1 ............................................................................................................. 1 1 1

xiii

Figure 4-22: COD concentration before and after air injection in location lift 1 ...... 113
Figure 4-23: COD concentrations decrease during air injection in lift 3 ................... 114

Figure 4-24: COD concentration vs. pH before and after air injection in sampling location

W111 — Lift 1 ............................................................................................................. 114
Figure 4-25: Acetone concentrations before air injection in lift 1 (9/2/2003) ........... 117
Figure 4- 26: Acetone concentrations after air injection in lift 1 (10/15/2004) ......... l 17
Figure 4-27 Fitting Monod equation to the experimental data for location G221 ..... 123

Figure 4-28 Monod equation based on the estimated parameters for location G221, T =

50°C,y=1.15 .............................................................................................................. 125
Figure 4-29: First order POCR for location G221 ..................................................... 128
Figure 4-30: First order POCR for two tests separated by 7 days of air injection ..... 129
Figure 4-31: Temperature during the two tests separated by 7 days of air inj ection..130

Figure 4-32: Volumetric moisture content during the two tests separated by 7 days of air
injection ...................................................................................................................... 131

Figure 4-33: Oxygen concentration before turning off the air injection .................... 132

Figure 4-34: POCR Difference between of tests 1 and 2 vs. their temperature difference
.................................................................................................................................... 133

Figure 4-35: Oxygen consumption rate at vent L6 (Date: 10/24/03) ......................... 134

Figure 4-36: Comparison between POCR and oxygen consumption rate in vents during
test 1 ........................................................................................................................... 135

Figure 4-37: Comparison between POCR and oxygen consumption rate in vents during
test 2 ........................................................................................................................... 135

Figure 4-38: Methane concentration as a function of time during startup in a sampling
location in lift 2 .......................................................................................................... 137

xiv

Figure 4-39: Methane concentration as a ﬁmction of time during startup in a sampling

location (G134) in lift 1 ............................................................................................. 138
Figure 4-40: Distribution of the rate of methane generation in Northern Oaks bioreactor
landﬁll during startup ................................................................................................. 139
Figure 4-41: Rate of increase in methane concentration (%lday) for each lift ......... 143
Figure 4-42: Methane concentration during start up of bioreactor landﬁll lift 4 ....... 145
Figure 4-43: Hydrogen concentration during and after lag time .............................. 147
Figure 4-44: Hydrogen concentration during start up of bioreactor landﬁll lift 4 ..... 148
Figure 4-45: Rate of increase in methane concentration (%lday) for each lift ......... 150
Figure 4—46: Methane regeneration after air injection was stopped in lift 2 .............. 155
Figure 4-47: Hydrogen concentration during and after lag time after air injection .13;

Figure 4-48: Hydrogen concentration change after air injection was stopped in lift 2....
.................................................................................................................................... 158

Figure 4-49: Methane concentration and waste temperature during the startup period for
lifts 1 and 2 ................................................................................................................. 161

Figure 4-50: Methane concentration and waste temperature during the startup period for
lifts 3 and 4 ................................................................................................................. 161

Figure 451: Rate of increase in methane concentration vs. waste temperature during

startup period ............................................................................................................. 162
Figure 4-52: COD vs. initial waste temperature in the landﬁll .................................. 162
Figure 4-53: Methane concentration after 400 days vs. COD ................................... 163
Figure 4-54: Rate of increase in methane concentration vs. pH ................................ 164

Figure 4-55: Rate of increase in methane concentration vs. waste temperature after air
injection ...................................................................................................................... 165

XV

Figure 4-56: Methane concentration after 62 days from stopping air injection for lifts 1, 2
and 3 and after 40 days from stopping air injection for lift 4 vs. waste temperature ......
.................................................................................................................................... 166

Figure 4-57: Rate of increase in percent methane after air injection vs. before air injection
.................................................................................................................................... 166

Figure 4-58: Comparison between the rate of methane concentration increase during
startup up and after stopping air injection .................................................................. 167

Figure 4-59: Comparison of the hydrogen concentration during lag time between during
startup up and after stopping air injection .................................................................. 168

Figure 4-60: Comparison of the hydrogen concentration after lag time between during
startup up and after stopping air injection .................................................................. 168

xvi

 

1. INTRODUCTION

Bioreactor landﬁlls are receiving considerable attention as an alternative to traditional
“dry tomb” landﬁlls. This is because bioreactor landﬁlls have several advantages over
traditional landﬁlls, including faster methane recovery (in case of anaerobic bioreactor
landﬁlls), additional space due to degradation of organic waste and settlement, and lower
post-closure maintenance [1]. Based on the electron acceptor conditions, bioreactor
landﬁlls can be operated in four different modes: aerobic, anaerobic, nitrifying, and a
combination of aerobic/anaerobic processes[2]. However, anaerobic mode of operation is

the most common [2, 3].

The ﬁrst bioreactor landﬁll was probably tested around 19805 [4]. An accurate count of
the number of bioreactor landﬁlls is seldom available due to the disagreement among the
practitioners about its deﬁnition. According to the United States Environmental
Protection Agency (US EPA) deﬁnition, “A bioreactor landﬁll operates to rapidly
transform and degrade organic waste. The increase in waste degradation and stabilization
is accomplished through the addition of liquid and air to enhance microbial processes
[4].” The Solid Waste Association of North America (SWANA) deﬁnes a bioreactor
landﬁll as “Any permitted subtitle D landﬁll or landﬁll cell where liquid and/or air, in
addition to leachate and landﬁll gas is injected in a controlled fashion into the waste mass
in order to accelerate or enhance biostabilization of the waste.” However, according to a
survey reported by SWANA in 2002, there were 20 full-scale demonstration bioreactor

landﬁlls in the US. [2].

Most ﬁeld scale anaerobic bioreactor landﬁlls focus on leachate recirculation as a major
variable [5-13]. Many other variables that may inﬂuence the extent of biodegradation of
waste in bioreactor landﬁlls are studied to a lesser extent. These variables include
alternative daily and intermediate covers and ﬁnal caps, gas extraction, compactness,
leachate seeps, addition of nutrients, preprocessing of waste, and heterogeneity of waste
[2]. Field scale studies focusing on temperature enhancement in bioreactor landﬁlls in
cold climates, oxygen consumption and heat generation rates, impact of aerobic start up
on methanogenesis, and measures of biostabilization end points have not yet been
reported. Heterogeneity which is always a concern with most measurements associated
with solid waste studies is also described in the context of the above parameters. The

following paragraphs describe some of these factors in more detail.

1.1. Temperature
Reported temperature in traditional landﬁlls is either in the mesophilic [14] or

therm0philic range [7, 15-17]. Mesophilic refers to a temperature range of 20 to 40 0C

[18, 19], and thermophilic refers to temperatures above 45 0C [20, 21]. In cold climate,

waste ﬁlled during winter may be at temperatures below freezing. At these temperatures,

only psychrophilic microorganisms are expected to be active. Psychrophilic
microorganisms have optimum temperature below 20 °C [20, 22]. If the temperature is

close to zero, even the psychrophilic microorganisms may work at much slower rates.
Therefore, low temperature can become a major issue for bioreactor landﬁlls started

and/or operated in cold climates. To our knowledge, no ﬁeld scale bioreactor landﬁll

study has been reported at temperatures below 10 °C. It is obvious that besides leachate

recirculation, enhancement in temperature may be required for bioreactor landﬁlls

operated in cold climates to obtain higher biodegradation rates.

1.2. Heat Generation Factor

As mentioned above, the increase in temperature due to air injection is a function of the
amount of heat produced per kg of oxygen consumed within the system and the amount
of heat lost ﬁom the system. The amount of heat produced per kg of oxygen consumed is
known as “heat generation factor”. Heat generation factor for bioreactor landﬁlls has not
been reported yet. Estimation of this factor combined with the heat loss is critical to

quantitatively link the increase in temperature increase to the rate of air supply.

1.3. Oxygen Consumption Rate

A cost effective approach to increase the temperature in landﬁlls is through aerobic
oxidation of waste. Enhanced air diffusion during waste placement [23] and forced
aeration after waste placement are the two main approaches that may enhance waste
temperature. Air injection is less intrusive to landﬁll operations when compared to
enhanced air diffusion. Air injection leads to creation of aerobic zones within the landﬁll
ensuing exothermic biodegradation process and increased temperatures. The amount of
heat produced per unit mass of organic matter degraded is at least 20-fold higher during

aerobic process than the heat produced during anaerobic process [24].

The design of air injection system is still in its infancy. Often a separate pipe network is
not necessary. The leachate and gas collection system may serve as part of an air
injection system, provided care is taken to insure ﬁre prevention and allow onset of
methanogenic activity. Currently, design of air injection system mostly amounts to
evaluating the amount of total air that should be provided at a given ﬂow rate. It is
evident that the rate of air supply will be a ﬁmction of the oxygen consumption rate
(OCR). Oxygen consumption rate may be deﬁned as the decrease in oxygen
concentration with time at a given location in a bioreactor landﬁll. The oxygen
consumption rate is a function of temperature and moisture. Oxygen consumption rate
expressed per unit mass of solid waste is also known as oxygen uptake rate. Ideally, only
the organic fraction of the solid waste should be considered in deﬁning oxygen uptake
rate. Data about the rate of oxygen consumption in landﬁlls is currently not available. In
composting, however, the effect of temperature on oxygen consumption has been
quantiﬁed using Arrhenius equation [25, 26] or by empirical relationships [24].
Arrhenius equation only considers temperature but some of the empirical equations also
include the effect of moisture [27]. To design an air injection system to enhance
temperature in bioreactor landﬁlls, it is essential to quantitatively relate the oxygen

consumption rate to temperature and moisture content.

1.4. Impact of Air Injection on Methanogenesis

Air injection will obviously impact the anaerobic processes leading to methane
production [28]. Methanogens are strict anaerobes and exhibit toxicity to small

concentrations of oxygen [29-31]. If intermittent air injection is to be used successfully

in enhancing temperature in anaerobic bioreactor landﬁlls, it is important to evaluate the
extent of impact on methanogenesis and the time of recovery for methanogenesis after air
injection is stopped. Recovery of methanogenesis after air injection is similar to the start
up phase of an anaerobic digestion process and air injection is akin to a perturbation to
the anaerobic process. Thus some measure of the anaerobic process may be helpful to
quantify the rate of recovery of methanogenesis. Besides methane production rate,
accumulation of hydrogen is a common indicator of upset anaerobic process. Elevated

levels of hydrogen may serve as an indicator of inhibited methanogenesis similar to those
reported for anaerobic digesters. Indeed, elevated H2 levels correlated with low pH and

decreased rate of methanogenesis in the only study (to our knowledge) reporting

hydrogen concentrations in solid waste samples from a landﬁll [32].

1.5. Heterogeneity

Heterogeneity in waste characteristics is one of the most difﬁcult problems encountered
when studying bioreactor landﬁlls. Often, aggregate parameters for the whole landﬁll are
the only measures of its performance. For example, the extent of biodegradation is
measured mainly by total gas produced or total leachate generation and its characteristics.
Sometimes, variability in the extent of biodegradation in landﬁlls is studied by
excavation of the waste followed by measurements in the laboratory for various
parameters, such as pH, lignin and cellulose content, biological methane potential,
volatile solids, and moisture content. Temporal measurements can not be made at the
same location using excavated samples. Alternative strategies, e.g., use of sampling ports

may be necessary to enable multiple measurements at the same location at least for

parameters that can me measured without destructive sampling. This study will also
present unique sampling strategies for leachate and gas sampling from within the landﬁlls

to characterize heterogeneity.

We have been working to establish a bioreactor landﬁll at the Northern Oaks Waste
Recycling Facility at Harrison, MI since August 2002. The cell construction took one

year during which the waste was exposed to temperatures below freezing due to the

winter season. During the ﬁlling period, ambient temperature ranged between -33°C to

+33°C (Figure 1-1) and the bioreactor landﬁll cell height increased to about 65 feet with
approximately 10 feet lifts. At the close of the waste placement, the temperature within

the waste was in the range between -5 °Cand 18 °C (Table 1-1) in the landﬁll. In lift 5

freezing conditions existed. Although ambient temperatures increased to 33°C during

summer months, the waste temperature remained low. Table 1-1 summarizes the
average waste temperature for each lift in the landﬁll on 9/5/03 (Day 420 from the start of
ﬁlling lift 1). Anaerobic digestion at such low temperatures is very slow if any, hence

increasing the landﬁll temperature is critical to enhance biodegradation.

 

 

 

 

 

 

 

 

Landﬁll height (ft)

 

 

 

 

 

 

 

 

Figure 1-1: Ambient Temperature during waste placement

Table 1-1: Temperature measured in the landﬁll on 5/9/03

 

 

 

 

 

 

 

 

 

 

 

Duration 0
after Temperature( C)
Lift Filling completion of
duration ﬁlling of the .
corresponding Average SD Max Mm
lift (months)

Jul 02-

1 Aug 02 13.7 12.6 1.1 14.7 10.5
Sep 02-

2 Oct 02 10.7 13.4 2.5 18.0 8.6
Nov 02-

3 Dec 02 9.4 9.5 4.1 16.2 3.7
Dec 02-

4 Jan03 8.4 11.1 3.7 14.7 4.4
Jan 03-

5 Feb 03 6.5 1.6 9.0 14.9 -4.8

 

 

1.6. Objectives

The goal of this research is to develop an engineered approach to enhance anaerobic
biodegradation of solid waste in bioreactor landﬁlls located in cold climates. Intermittent
air injection system is the method selected for increasing the temperature. To obtain
design parameters needed to operate bioreactor landﬁlls in cold climates, the following

speciﬁc objectives are set.

Objective 1: To establish and quantify relationship between amount of heat produced
(kJ) per unit mass of oxygen supplied (kg), and to characterize the performance of air

injection with respect to temperature and leachate characteristics.

Hypothesis: Heat generation factor determined in laboratory conditions for single

substrates may differ from that obtained for solid waste under ﬁeld conditions.

Objective 2: To establish the spatial variability in point oxygen consumption rate
(decrease in oxygen concentration with time at a given location in a bioreactor landﬁll),
and study the effect of temperature, oxygen concentration and moisture content on point

oxygen consumption rate.

Hypothesis: Point oxygen consumption rate is a function of moisture content and
temperature. The increase in moisture content will inhibit POCR however the increase in

temperature will enhance POCR.

Objective 3: To quantify the performance of anaerobic digestion at startup and following
air injection with respect to methanogenesis establishment, using methane and hydrogen

concentrations as indicators for anaerobic digestion process.

Hypothesis: Air injection will affect methanogenic activity negatively however this affect
can be altered by improved conditions of enhanced temperatures. Thus air injection will
enhance the methanogenic activity after the aeration process is ceased compared to before

air injection.

2. LITERATURE REVIEW

Although a large number of studies exist on traditional landﬁlls, studies related to
bioreactor landﬁlls are relatively few. Hence the following paragraphs extract
information related to i) temperature, ii) oxygen uptake rate, iii) heat generation factor,
and iv) potential impact of intermittent aeration on methanogenesis in bioreactor landﬁlls

from published literature on bioreactor landﬁlls a well as traditional landﬁlls.

2.1. Temperature
Temperatures reported in landﬁlls at the time of placement generally range between 10 to
20°C [15 , 16]. This temperature may be lower in colder climates but no published data is

available for landﬁll temperatures in cold climates, especially during initial stages of
ﬁlling. Because the focus of this project was to enhance the temperature of waste placed
in bioreactor landﬁlls in cold climates, studies focusing on the behavior of temperature
with time and various operations are reviewed. Two main strategies, namely air diffusion
during waste placement and active air injection, are reviewed below for their

effectiveness in enhancing landﬁll temperature through aerobic degradation.

2.2. Enhanced Air diffusion
Enhanced air diffusion can be achieved by either increasing the porosity of waste or
lengthening the waste placement. Under normal ﬁlling operations, a 10-15 °C increase in

temperature can be achieved by enhanced air diffusion [33, 34]. Lowering the refuse

density and adding wood chips or aerobically stabilized waste are the two main

10

approaches to increasing porosity [35]. Lengthening the exposure time to air was used in
one case with approximately 25°C change in temperature [15]. Ambient temperatures do

not signiﬁcantly impact the waste temperature after the landﬁll is covered. This has been
observed in many cases [36, 37] and is due to the elimination of air diffusion and thermal

insulation by the waste.

Lifts closer to the surface of the landﬁll generally have lower temperatures compared to
lifts that are closer to the base [15, 17, 38, 39]. However, exceptions to this may exist
due to seasonal variations during waste ﬁlling. Waste ﬁlled at higher ambient
temperatures will have steady state temperatures that are higher than those ﬁlled at lower
temperatures. In one case, it was observed that temperature in various lifts may slowly
approach a constant value for each lift in the absence of leachate recirculation [37]. This
steady state temperature may depend upon the initial temperature at which the waste was

ﬁlled. For example, at Yolo County bioreactor landﬁll, CA, lift one or the deeper layer
was ﬁlled at ambient temperature of around 15°C, which increased to about 25°C during
the aerobic phase. Ambient temperature during ﬁlling the second and third lifts was

about 25°C, which rose to about 50°C immediately after ﬁlling due to aerobic oxidation

but subsequently decreased to about 31°C [37].

Moreover, it was observed from ﬁgures published in [37] that temperatures in waste
between consecutive lifts, if different from each other, might stay different for long

periods of times. This can be explained by the minimal heat conduction occurring in the

waste matrix due to its low thermal conductivity 0.1i0.04 Wm'lK.1 [15]. Such an

11

insulation property is very close to that of asbestos which is known for its ability to

insulate heat and has a thermal conductivity is 0.05-0.15 Wm'lK'l [40, 41].

Although temperature increase can be achieved by diffusion, it is not preferred in practice
because ﬁlling must be continuous; leaving the air space unﬁlled for air diffusion is not
an option. Air injection is a better option to achieve temperature increase because it does

not interfere with the landﬁll operations and does not delay the utilization of air space.

2.2.1. Air injection

Air injection is used to change the anaerobic conditions in a landﬁll into aerobic state. If
aeration is adopted at the whole bioreactor landﬁll for very long periods, the cell will
obviously turn into an aerobic bioreactor landﬁll. Partial aeration of the landﬁll results in
hybrid landﬁlls; in hybrid landﬁlls the upper layers of waste are aerated, while bottom
layers are operated under anaerobic conditions [4]. There are few reported aerobic
bioreactor landﬁll demonstration projects [42-47] and two hybrid bioreactor landﬁlls [16,

48]. Air ﬂow rate reported in these bioreactor landﬁlls varies considerably, ranging from

0.2 to 2.9 m3air/m3 waste per day (Table 2-1). Such ﬂow rates are sufﬁcient to cause an
increase in temperature (Figure 2-6). The rate of temperature increase ranged from 0.04

°C/day to 1.3 °C/day (Table 2-1). Occasionally, much faster rates of increase have also

been observed, e.g., an increase of 20 °C in 5 days was reported due to air injection in a

landﬁll in Korea [46]. This is the fastest rate of increase in temperature reported for

landﬁlls. The total increase in temperature can also vary signiﬁcantly. For example at

12

Live Oak landﬁll (Georgia), the average temperature reported was 37.8 °C but ‘hot

pockets’ of up to 71 °C were also reported [42]. The increase in temperature is generally

limited to the duration of aeration. In the hybrid Metro Bioreactor Landﬁll (Wisconsin),
as soon as the air injection ceased, the temperature started to decrease [16]. However, at
steady state, temperatures stabilize at higher temperatures compared to those before
aeration. This pattern of changes and stabilization at higher temperature is similar to that

observed during start up of landﬁlls as a result of air diffusion.

Table 2-1: Air ﬂow rate and temperature change as reported in aerobic bioreactor

 

 

 

 

 

 

 

landﬁlls
Air 12°"; t Rate if Waste Flow
rate m emper a ure Volume Rate Landﬁll Name
air/m3 increase Location ’ Reference
0 (m3 ) (m3/ min)
waste day) ( C/day)
Live oak
2.9 Not reported 49,000 100 landﬁll, [45]
Georgia
Live Oak
0.2 0.3 53,500 6.1 Landﬁll, [42, 49]
Georgia
0 3 Columbia
1.8 ° 45,200 56 county landﬁll, [43-45]
Georgia
Williamson
28.3 x 3 = County
0.7 0.04 185,000 84.9 bioreac tor, [47]
Texas
4 Not
590,000 reported Seoul, Korea [46]
1.3 Not Not Metro Landﬁll, [16]
reported reported Wisconsin

 

 

 

l3

2.2.2. Radius of influence

Air injection into landﬁlls is a relatively new practice. Hence, parameters such as the
extent of dissemination of oxygen and nitrogen (described by the radius of inﬂuence)
during air injection through vertical wells or horizontal pipes is rarely estimated. The
radius of inﬂuence with respect to pressureiof an injection well may be deﬁned as the
distance from the injection point at which a pressure change of 0.1 inch of water, or about
10% of the injection pressure is observed [46]. Injection rate and pressure distribution
are important for the design of the air injection and distribution systems. As the radius of
inﬂuence increases a less dense air injection system network is required. The radius of
inﬂuence with respect to oxygen is reported to be less than that given by pressure
changes because oxygen is consumed during dissemination [46]. This radius of oxygen
is the distance from the injection line to the point where the oxygen concentration
becomes negligible [46], and is an important parameter for designing air injection system
because it determines the maximum allowable distance for the placement of pipes used
for air injection. The difference between the radius of oxygen and nitrogen is determined
by the kinetics of oxygen consumption. A faster rate of oxygen consumption results in
bigger difference between pressure and oxygen radii of inﬂuence. Both radii are a
function of the rate at which air is being injected and characteristics of the waste.

Determining this rate under ﬁeld conditions is critical for designing air injection systems.

2.3. Heat Generation Factor

Heat is produced by aerobic biological activity [50]. The amount of heat generated per

unit time is known as the heat generation rate. This rate is proportional to the oxygen

14

uptake rate as well as the carbon dioxide production rate and the biomass concentration
[50—52]. The proportionality constant linking the heat generation rate to the oxygen
uptake rate is called the heat generation factor. The heat generation factor is sometimes
also referred to as heat yield on oxygen. To measure the heat generation factor, a heat
balance around the bioreactor is conducted using a heat balance, i.e., by stating that
accumulation of heat equals the heat generation in the system minus the heat loss from
the system. Both heat accumulation and heat loss can be quantiﬁed [51, 53-57].
Substituting these values in the heat balance equation and solving the resulting equation

yields the heat generation factor.

The heat generation factor can be measured in the laboratory either by dynamic or
continuous calorimetry [52]. The dynamic method allows the temperature to rise in the
bioreactor, hence accumulation of heat takes place with time [51, 53-56]. The slope of
curve in which temperature is plotted against time is used to calculate the accumulation
of heat. The loss of heat in the system is measured and substituted into the heat balance
equation. This equation is then solved to obtain the amount of heat produced at several
time points. The slope of the calculated amount of heat produced vs. the oxygen uptake
is the heat generation factor. Continuous method maintains a constant temperature by
heating or cooling the reactor [54, 55, 57]. In the continuous method, the accumulation
of heat is zero because the temperature is constant. Therefore, heat generation equals the
heat loss in the system. In laboratory scale reactors, it is feasible to cool and/or heat the
system. However it is not feasible to maintain constant temperatures in large-scale or
ﬁeld-scale reactors. Therefore, the dynamic method is the method of choice under ﬁeld

conditions. It is also acceptable because heat generation factors measured by both

15

methods are comparable. Turker [55] obtained a heat generation factor for molasses of
444 kJ/mole Oz and 431 kJ/mole 02 using continuous and dynamic methods, respectively

[55].

There is lack of data for heat generation factor for solid waste. However, studies related
to heat generation factor for single substrates including glucose, molasses, soybean meal,
ethanol, lactose, and acetate are plentiful [24, 51, 52, 55, 56]. Turker reported the range

of heat generation factors for the above single substrates in the range of 397 to 543

kJ/mole 02 using laboratory bioreactors [55]. At the average heat generation factor for

the above substrates was 456:38 kJ/mole Oz indicating a relatively small variability [55].

In another study, the slope of heat generation rate vs. oxygen consumption rate for

different single microbial species grown on different single substrates gave a value of 444

kJ/mole Oz [50].

Values for heat generation for synthetic solid waste are lower (304 kJ/mole Oz [24]) than

for single substrates. Only one study considered the use of large scale bioreactors as
calorimeters to collect data for heat generation factor [55]. However, in most instances,

heat generation factor obtained for single substrates have been routinely used for complex

substrates including solid waste. For example, a heat generation factor of 443 kJ/mole 02

was used in composting operations [58] and a value of 460 kJ/mole 02 was used for a

landﬁll [23]. It is evident that determination of heat generation factor for solid waste

preferably in situ is critical to obtain a better understanding of temperature distribution in

16

 

bioreactor landﬁlls and predict the amount of oxygen required to increase landﬁll

temperature.

2.4. Oxygen Consumption Rate

Oxygen consumption and /or carbon dioxide production have been used extensively to
indicate the status of aerobic degradation [25, 53, 59-72]. The most common way of

expressing the oxygen consumption rate is by using the amount of oxygen consumed

(weight of oxygen consumed per hour (g Oz/hl') to the mass of solid waste in a control

volume (kg)) [72]. The mass of solid waste obviously requires further description about
the type of waste. Total organic matter [62] , volatile solids [61], and volume [55] of
solid waste have also been used instead of total mass (Table 2-2). The oxygen
consumption rate can be determined in the laboratory by respirometry either under
dynamic [53, 59-65] or static conditions. However, the determination of the oxygen
consumption rate by respirometry requires high levels of oxygen in the sample and outlet
gas (above 14%). In bioreactor landﬁlls, the concentration of oxygen in the efﬂuent air
in response to air injection might not reach such high levels. Therefore, the respirometric
method can not be applied under ﬁled conditions. It is possible, however, to determine
the rate of oxygen consumption at a given point in the landﬁll cell. Determining this rate
under ﬁeld conditions is necessary because rates determined under laboratory conditions
are not reﬂective of the ﬁeld rates. For example, the OCR reported for full-scale plants

were almost twice as much as those determined in a laboratory scale system [73].

Oxygen uptake rates reported in literature vary in the range of 1.0-1.5 mg g'1 hr.1 for an

active composting process; above 1.5 mg g'1 hr'l for very active (considered unstable);

17

and up to 3.5 mg g'1 hr.l for extremely high activity [72]. Oxygen consumption rate for

landﬁlls have not been determined before, therefore only values for composting of solid

waste is reported.

2.4.1. Effect of process variables on aerobic degradation

Several studies report correlations between oxygen consumption, carbon dioxide
production, and aerobic biodegradation rates with other process variables such as
temperature [24, 25, 72, 74-77], dissolved oxygen concentration /gas phase oxygen
concentration [25, 78], maturation of waste [62, 73], air ﬁlled porosity [79, 80], and
moisture content [72, 81]. The inﬂuence of these parameters on oxygen uptake rate
varies. Kulcu [27] summarized that a minimum concentration of 5% in the pore space is
required for composting piles. Temperature is an important factor for oxygen uptake,
however it less inﬂuential than moisture [72]. A lag phase occurs at the onset of aerobic
degradation regardless of moisture content. This lag phase decreases as the temperature
increases [72]. OCR depends on the waste moisture content [72]. Moisture contents of
50, 60, and 70 % did not signiﬁcantly impact the OCR. However, moisture contents of
30, 40, and ~50% signiﬁcantly impacted the OCR. A minimum moisture content of 50 %
was required for active composting, with no upper limit in the moisture content at the
ranges studied (30-70 %) [72]. Similarly, air ﬁlled porosity is also known to affect

microbial kinetics. Maximum OCR were observed at air ﬁlled porosity of 25-30 % [80].

As described above, several factors affect the oxygen consumption rate. However the

effect of temperature is most commonly studied. In the composting literature, it is

18

observed that temperatures below 20 °C severely slowed down the composting process
[72]. The decomposition rate in compositing operations are also reported to slow down

under high temperature conditions. Temperatures above 60 °C were reported to reduce
the microbial activity and these above 82 oC severely hindered the process [72]. The

highest OCR in composting is at temperature range of 52 to 60 °C [72]. Within this

optimum composting temperature range, the oxygen uptake rate increased with an

increase in temperature.

Several empirical and mechanistic equations are available to quantify the effect of
temperature on oxygen uptake rate. Among the mechanistic equations, the Arrhenius
equation is the most common [25, 26]. It can be applied to the individual temperature
ranges in which microbial activity takes place (i.e., psychrophilic, mesophilic, and
thermophilic) or to the entire range [26]. Often empirical equations were found more

appropriate as they can better describe the effect of temperature [24].

19

Table 2-2: OCR reported for solid waste.

 

 

 

 

 

 

 

 

OCR* Process Reference
0.11-4.73( g 02/ (kg TOM Organic fraction of municipal solid waste [62]
hr)) (OFMSW) composting operation
0.130-0.930 (g Oz/(kg vs [61]
hr»
10'5-10'3 (mol oz /( m3
sec)) Solid waste laboratory column fed by [15]
0806-8064 (11 (m3 hr)) diffusion only
0.0036036 (mol/ (m3 hr))
Mechanical-biological end composting
0500 (g 02 / (kg VS hr)) process (pilot scale) -Active phase [73]
Mechanical-biological end composting
0280 (g 02 /( kg VS hr)) process (full- scale) — Active phase [73]
1-0 (mg/(8110) [72]

 

‘Note that OCR is reported in the original units which may be different for each reference.

2.5. Anaerobic Digestion of Solid Waste

Anaerobic digestion is characterized by greater degree of metabolic specialization in

comparison to that of aerobic degradation [28]. The ﬁrst step in the anaerobic digestion

of complex waste is hydrolysis in which particulates substrates such as lipids,

polysaccharides, proteins and fats are hydrolyzed and fermented into soluble organic

materials [82-86].

The second step is the conversion of these soluble compounds by

acidogens into volatile fatty acids (propionate, butyrate etc.), carbon dioxide, and

hydrogen [85]. The third step is the conversion of these fatty acids by acetogens into

acetic acid and hydrogen. The last step of anaerobic digestion is methanogenesis, in

20

 

which hydrogen and carbon dioxide are converted to methane by COz-reducing

methanogens or acetate is cleaved to methane and C02 by aceticlastic methanogens.

Approximately 75 % of the methane produced is a result of acetate conversion into
methane and carbon dioxide [28, 86, 87]. Hydrogen utilizers are responsible for

producing the rest of the methane [86, 88, 89]

In line with the anaerobic digestion process, anaerobic decomposition of solid waste is
marked by an increase in chemical oxygen demand (COD) and volatile fatty acids (VFA),
and a decrease in pH [90-93]. This stage is usually followed by an exponential increase
in methane generation, pH increase, and a decrease in VFA and COD. At the end,
methane production starts to decrease and COD stabilizes at low concentrations. The
only difference under ﬁeld conditions may be in the rates at which the above processes

occur [94].

The start-up of anaerobic digestion is considered the most critical step in the operation of
anaerobic digester [95]. It is characterized by low abundance of methanogens and
acetogens and high abundance of acidogens growing at a fast rate [96]. This imbalance
leads to the accumulation of hydrogen and VFA (and therefore low pH). These
intermediates further inhibit acetogens and methanogens. Methanogenesis may further be
affected by lower temperatures, leading to a very slow start up of the process under cold

climates [97].

21

2.5.1. Effect of temperature on anaerobic digestion

Mesophilic operation of anaerobic digesters is most common followed by the
thermophilic process. Landﬁll temperatures above 30 °C have been reported to be
favorable for the anaerobic digestion of solid waste [35]. The optimum temperature range
for refuse degradation in the laboratory was found to be 30-35 °C [35]. Engineered
anaerobic digesters are probably nonexistent under psychrophilic conditions (10 to 20
°C). Microbial degradation rates are expected to be very slow under psychrophilic

conditions. When the temperature decreases from mesophilic to psychrophilic range,

methane production decreases [98]. This is a known to be a problem during winter [83].
The start up of anaerobic reactors at low temperatures (15 °C and 25 °C) is problematic

even under laboratory conditions [97]. Besides affecting the startup of a bioreactor,
psychrophilic temperatures also impact the structure of the microbial community. For
example under psychrophilic conditions, acetoclastic methanogens are reported to be the
main microorganisms for methane production [99, 100]. Hydrogen consumers are
hindered more at such temperatures leading to accumulation of hydrogen. Homoacetate
bacteria strongly compete with methanogens and might be considered “the most typical
participants of the anaerobic decomposition community in psychrophilic conditions”

[97].

Thermophilic anaerobic digestion process has shorter startup time and higher gas
production rates compared to the mesophilic process [95]. Growth rates of methanogens

under thermophilic conditions are higher than those observed under mesophilic

22

conditions. As the temperature increases to the hyper-thermophilic range (i.e., above 80
°C), hydrogen consuming methanogens become more active, often exceeding the rate of
hydrogen producing syntrophs [101]. The optimal temperatures for methane generation

from acetate are 55 °C, for hydrolytic and fermentative bacteria are 55 to 70 °C, for
hydrogen utilizing methanogens 65 oC, and for acetate, forrnate and butyrate consuming

populations 60 oC. Thus, in thermophilic anaerobic digestion process, different groups

of microorganisms require different optimal temperature, complicating the optimal

performance [82].

2.5.2. Inhibition of methanogenesis due to low pH and high VFA

The pH in an anaerobic digestion process is an indicator of imbalance and stability in
anaerobic digestion system. The vulnerability of reactors to pH changes is due to
variability in buffering capacity (V FA and bicarbonate) [98]. High VFA concentrations
are associated with low pH values. The methanogens are inhibited by the unionized acid
concentration, leading to a decrease in methane production [88, 102, 103]. Inhibition
during the anaerobic digestion of solid waste due to VFA is well documented [91].
However, the concentration of VFA at which inhibition occurs in solid waste is much
higher compared to anaerobic digesters [91]. Optimum pH for hydrolysis is slightly
acidic (~pH 5) and inhibition occurs at pH of 4.0 [85]. Optimal pH for methanogens is
between 7 and 7.2. Inhibition occurs at pH values below 6.2 or above 7.8 [20, 87, 104].
Butyrate degradation is optimum in the pH range of 6.8 to 7.3, with inhibition at pH 6.0

to 6.4 [105].

23

2.5.3. Inhibition of methanogenesis due to oxygen

Methanogens are strict anaerobes, hence exposure to air is detrimental to the anaerobic
process [29]. Inhibition constants (the concentration at which 50% inhibition occur) for
oxygen are at very low concentrations 0.0001-0.0005 gﬁr L'1 [31]. However, it is also
reported that granular sludge survived up to 18 days when exposed to air and substrate
was available [29]. High dissolved oxygen concentration (8.1 ppm) in the inﬂuent
wastewater was inhibitory to acetogens and methanogens [29]. A slight decrease on
acetate and propionate utilization rate was found at 8.1 mg L'1 dissolved oxygen when
operated for 2 months. This rate declined further over prolonged operation. The
structure of the bioﬁlm was able to protect the strict anaerobes from oxygen toxicity
when exposed for one month to oxygen. The tolerance of methanogens to oxygen was
attributed to their presence in structured granules, protected by facultative bacteria and
aerobes at the outer parts limiting the diffusion of oxygen. Longer aeration periods leads
to the destruction of these aggregates, leading to easier diffusion of oxygen and

penetration of oxygen to deeper layers [29].

2.5.4. Hydrogen and anaerobic digestion process

Hydrogen is produced (together with acetate) by fermenters and syntrophs as a by
product during anaerobic digestion process. Acetate is produced by three paths:
Approximately 30% of the acetate is produced by the fermenters. Sixty percent of the
total acetate is produced by the conversion of butyrate, propionate, valerate, isovalerate,
and isobutyrate to acetate by acetogens (syntrophs). The rest is formed from forrnate,

hydrogen and carbon dioxide by the homacetogenic bacteria [106]. Acetate production

24

by syntrophs is inhibited by elevated levels of hydrogen concentrations. A theoretical
value of 100 ppm of hydrogen in the gas phase is shown to thermodynamically inhibit the
syntrophs [92]. In addition methanogens that utilize acetate at thermophilic conditions
such as Methanosarcina therrnophila TM-l are also inhibited by hydrogen concentrations
above 250 Pa (~2500 ppm) [107]. The percent inhibition at 76, 101, 253, 507, 1010 and
2030 Pa is 0%, 5% 13%, 31%, 82%, and 100%, respectively. This pressure inhibited
syntrophic butyrate degrading cultures, but at lower level than that of acetate utilizing
methanogens [107]. Syntrophs particularly butyrate degraders at hydrogen pressure of 76,
101, 253, 507, 1010, 2030 and 3040 Pa were inhibited by 0 %, 3 %, l7 %, 32 %, 52 %,
75 %, and 100 %, respectively. It seems at lower hydrogen concentrations (up to 507
Pa) the inhibition of syntrophs is slightly worse than those of acetate utilizing
methanogens, however as the hydrogen pressure further increases syntrophs are less
inhibited. These values for inhibition of syntrophs are at much higher levels than 100
ppm. This implies that the methanogens that utilize acetate, known to produce more
than 70 % of methane in anaerobic digestion, are more affected than syntrophs, indicating
that the anaerobic digestion under such conditions will further be inhibited, due to acetate

accumulation and reduction in pH.

The level of inhibition by hydrogen has been reported to vary signiﬁcantly between
suspended and attached grth anaerobic digestion. A low dissolved hydrogen
concentration in anaerobic digestion is an indication of a well functioning anaerobic
process. Digesters with upsets experience elevated levels of hydrogen [86]. It is reported
to be a useful indicator of process upsets because the effect on hydrogen is more rapid

than on methane [93, 104]. Others state that hydrogen is not a good indicator because of

25

mass transfer limitations [101, 107] and fast dynamic behavior [97, 108]. Its relevance
as an indicator is less documented in attached growth systems. Some researchers believe
that aggregates and bioﬁlrns provide an efﬁcient mechanism for hydrogen delivery and
its removal [101]. In granular sludge microcolonies have been observed between
syntrophs such as propionate or butyrate degraders and hydrogen utilizing methanogens.
Destruction of these aggregates lead to a decreased rate of activity for the syntrophic
microorganisms [102]. The proposed structure of methanogens and syntrophs in the
aggregate includes acidogenic bacteria in the outer layer and hydrogen utilizing bacteria
in the inner layer. Within such structures tolerance to inhibitory concentrations of

substrates is quite likely [102].

26

3. SITE DESCRIPTION AND METHODS

3.1. Site Description and Bioreactor Landfill Cell Construction

The bioreactor landﬁll is located at Northern Oaks landﬁll and Recycling Facility-
Harrison, Michigan which opened in December 1992. The facility, occupies 320 acres,
of which 76 acres is designated for active use. Nearly 75 % of the leachate collected on

site is recirculated either to the bioreactor landﬁll cell or to the working face of other cells

[109].

The bioreactor landﬁll cell (the subject matter of this study) is 60 feet deep, has 1.2 acres
footprint and contains 71,000 cubic yards of municipal solid waste (Figure 3-1). This cell
was ﬁlled between 7/12/2002 and 3/4/2003 and ﬁnal cap in place by 6/25/2003. The total
volume of cover soil used was 17,000 yd3 which is approximately 19 % of the landﬁll
volume. The cell includes horizontal leachate recirculation trenches, horizontal leachate
recirculation lines placed on geocomposite, horizontal gas extraction lines, and settlement
proﬁlers. The landﬁll cell was constructed in six 10-ft lifts referred to as lifts l to 6,
starting from the bottom. At the top of lift 1 leachate recirculation lines, gas extraction
lines, sampling ports, and probes were placed. The leachate distribution lines and ports
were placed at the top of lifts 1 to 5. Gas extraction lines were placed in the middle of
lifts 3, 4, 5, and 6, i.e., at 25, 35, 45, and 55 ft from the bottom, and at the top of lifts 1
and 2 at 10 and 20 ft from the bottom. The waste was covered daily with soil according
to the regular operation of the landﬁll at Northern Oaks. The gas extraction lines were
connected to a ﬂare through a gas header. Leachate from the bioreactor cell was

collected through drainage system constructed above the liner and connected to a sump.

27

The sump was equipped with a level controlled pump to meet the regulatory requirements
of 1 ft of leachate head above the liner. Table 3-1 lists the events of construction and

placement of monitoring devices at the bioreactor cell in a chronological order.

 

 

 

 

Figure 3-1: Bioreactor landﬁll cell at Northern Oaks Recycle and Disposal Facility. A.
The layout of the bioreactor landﬁll cell, B. View of the cell from the north during ﬁlling,
C. View of the cell from north after ﬁlling was completed.

3.2. Monitoring Devices

Monitoring devices at the bioreactor landﬁll site included leachate and gas sampling
ports, moisture and temperature probes, weather monitoring station, and settlement
proﬁlers. A total of 48 ports for gas and leachate sampling were placed within the
landﬁll in a three-dimensional grid. Moisture and temperature probes were installed next
to the sampling ports. The distribution of these monitoring locations was as follows: 19
in lift 1, 13 in lift 2, 6 each in lift 3 and 4, and 4 in lift 5 (Figure 3-2). A perforated 1
inch diameter and 3 inch long PVC pipe connected to a long tubing extended outside the
cell to be used as a gas sampling port (Figure 3-3). The leachate sampling port consisted

of a 14 L plastic bucket, with a perforated manifold that was placed inside the

28

Table 3-1: Chronological order of ﬁlling the Northern Oaks Bioreactor Landﬁll

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Start End
Description
Jate Day“ Date Day
ﬂZ/ZOOZ 0 8/1/2002 20 Fill Lift 1
ﬁ4/2002 12 7/24/2002 12 Install weather station
\8/24/2002 40 9/13/2002 63 Install monitoring devices on Lift 1
M/ZOOZ 82 10/8/2002 88 Install leachate recirculation and gas extraction lines on Lift 1
M002 89 10/28/2002 108 Fill Lift 2
M2002 111 11/8/2002 119 Install monitoring devices on Lift 2
M002 122 11/11/2002 122 Install leachate recirculation lines on Lift 2
M2002 123 11/22/2002 133 Fill Lift 3 (First 5 ft)
M002 136 11/25/2002 136 Install gas extraction lines on Lift 2.5
11/2 6/2002 137 12/6/2002 147 Fill Lift 3 (Second 5 ft)
12/ l 4/ 2002 155 12/16/2002 157 Install monitoring devices on Lift 3
12/ l 7/ 2002 158 12/17/2002 158 Install leachate recirculation lines on Lift 3
12/1 7/2002 158 12/26/2002 167 Fill Lift 4 (First 5 ft)
12/27/ 2002 168 12/27/2002 168 Install gas extraction lines on Lift 3.5
12/27/2002 168 1/7/2003 179 Fill Lift 4 (Second 5 ft)

1/24/ 2003 196 1/26/2003 198 Install monitoring devices on Lift 4
_1/2\8/‘2003 200 1/28/2003 200 Install leachate recirculation lines on Lift 4
$842003 200 2/4/2003 207 Fill Lift 5 (First 5 a)
£003 208 2/5/2003 208 Install gas extraction lines on Lift 4.5
M03 208 2/13/2003 216 Fill Lift 5 (Second 5 ft)

_2/1\3/2003 216 2/14/2003 217 Install monitoring devices on Lift 5

-2/1\4/2003 217 2/14/2003 217 Install leachate recirculation lines on Lift 5
M003 220 2/19/2003 222 Fill Lift 6 (First 5 ft)

£2003 223 2/20/2003 223 Install gas extraction lines on Lift 5.5

#244003 227 2/26/2003 229 Fill Lift 6 (Second 5 ft)

£10003 230 2/27/2003 230 Install leachate recirculation lines on Lift 6
_3/{/2003 234 3/4/2003 235 Fill Final Lift

Mom 306 5/14/2003 306 Install gas ports and TDT in cap below geomembrane
Mom 339 6/20/2003 343 Install geomembrane cap

__6/1&2003 341 6/25/2003 348 Covering geomembrane with soil

£2003 396 8/12/2003 396 Install gas ports and TDT in cap above geomembrane
MIL/2003 431 9/16/2003 431 Install three ﬂux chambers

 

29

 

     

m

nu

B

n

.m

m

m

m.
S
elvO
mp,
u T,
///ﬂw

;’ [l
//

//

  

 

 

 

 

 

     
        

 

   
 

s l 2 3 4 5
d r m ﬁ ﬁ ﬁ ﬁ h
n r 0 on .I :I .1 :l .1
e 0 m a L L L L L
m... e m .m .m m m .m
S

o A x n_ x

R 2 t
sass; L new...” m I
$53.. . Essex... L I
n

$Nw$ $3 .3.
. 4. »$ $3 «3
, £33,: «.25

                

tor landﬁll at NORDF

Figure 3-2 Layout of bio

30

 

l': ‘2' w 1:1, it'. .......

         

      

:x fOtZ' (in-I .4

ltuj-Z'.
' 3,;

t
a . x..

I

  

m; .. .»:t.c:;l:ztzzl:tISiulurtaW

Imam-

Figure 3-3: Gas sampling port
bucket, and with a long tube that extended to outside the cell for sampling. The bucket
was covered with a perforated plate to prevent large slid objects from entering the bucket

(Figure 3-4).

Two types of moisture probes were used: Time Domain Reﬂectometry (TDR; Automata,
Inc.) and Time Domain Transmissometry (TDT; GroPoint soil moisture sensor). Both
had the capability to measure temperature. Figures 3-5 and 3-6 show these probes. TDR
probes were placed at all the 48 locations mentioned above and TDT probes were placed
below the geomembrane cap and within the topsoil in intermediate cover. A weather
station (Campbell Scientiﬁc, Inc.) located at the east side of the cell (Figure 3-7),
provided continuous measurement of air and soil temperature, and precipitation.
Temperature and moisture probes and weather station were connected to a data logger
with multiplexer and remote data collection capabilities. Leachate ﬂow to the sump was

monitored by a ﬂow meter (model 52 EP125, EPG companies Inc.), which operates in the

31

range of 5.7-103.4 gpm. The total gas produced from the landﬁll was measured at the

header using a gas ﬂow meter (model TA2, Magnetrol International, IL).

 

 

 

 

Figure 3-5: Automata TDR moisture probe and hand held reader

32

 

Figure 3-6: Gro.point TDT moisture probe

 

 

 

 

 

Figure 3-7: Relative location of the weather station on site (This picture was taken from
east side of the bioreactor landﬁll)

3.2.1. Automated and onsite measurements

Online measurements included moisture, temperature, weather related data (wind speed
and direction, rainfall, snow accumulation, etc.), and leachate and gas ﬂow. In addition,
gas composition and temperature were measured onsite at the venting and extraction lines
using a gas analyzer (Gem-meter 2000). Pressure was also measured in the dry gas
sampling tubes using a manometer (Extech Instruments, NY), except in those tubes that

were full with leachate.

33

3.2.2. Gas and leachate sampling protocol for off-line analysis

Gas and leachate samples were collected using a peristaltic pump with a capacity to pump
gas and leachate at 150 ml min'1 and 300 ml min”, respectively. The pump was
connected to the sampling ports described above at the time of each sampling event.
Sampling tubes were purged for 15 minutes (3 times the volume of the longest tube)
before taking samples. Leachate sampling-buckets were emptied 7 to 14 days in

advance before each leachate sampling event.

Serum bottles (70 ml total volume) were prepared as follows for gas sampling: bottles
were completely ﬁlled with acidiﬁed water (pH < 2 using concentrated hydrochloric acid)
and sealed with butyl rubber stoppers and aluminum crimp caps. Gas samples were
collected by water displacement method. During winter months the acidiﬁed solution
also contained 20 % sodium chloride to prevent freezing. Bottles were stored at room

temperature until the time of analysis.

3.3. Analytical Methods

Gas samples were analyzed for oxygen, nitrogen, carbon dioxide, methane and hydrogen.
A l-ml gastight syringe (Hamilton) was used to take a sample from the headspace of the
gas sampling bottles. Leachate samples collected from the sampling ports were analyzed
for pH, electrical conductivity (EC), chemical oxygen demand (COD), and volatile
organic compounds (V DC). Protocols for measuring the above parameters are brieﬂy

summarized below.

34

3.3.1. Hydrogen

Hydrogen was analyzed by injecting one mL of headspace sample in a reduction gas

analyzer (RGA3, Trace Analytical) equipped with Spherocarb MoleSieve column 60/80;
31 1/4" x1/8" column at 53 0C, with the detector temperature set at 265 °C. The

retention time of hydrogen was 0.2 minutes using Nitrogen as a carrier gas with ﬂow rate
18-20 mL / min. The peaks were integrated using Turbochrom software (Perkin Elmer,

MA).

3.3.2. Oxygen, nitrogen, carbon dioxide and methane

Oxygen, nitrogen, carbon dioxide, and methane were analyzed using Gas
Chromatography HP5890 equipped with a thermal conductivity detector (TCD). Column
used was Supelco 60/ 80 Carboxen — 1000 and 15 ft 1/8 in outer diameter. Carrier gas

was set to a ﬂow rate of 40 ml/min. The temperature for the injector and detector were

set at 110 °C and 200 °C, respectively. Oven temperature was set at 45 °C for 5 minutes

as isothermal then increased at a rate of 40 °C/min to ﬁnal temperature of 200 °C. This

ﬁnal temperature was kept for 2 min before it was allowed to cool down to the initial

temperature. All peaks were quantiﬁed with Turbochrom software.

3.3.3. Electrical conductivity, pH, COD, VOC

Leachate COD was measured using either 0-1500 ppm or 0-15,000 ppm range

commercially available kits (HACH Company, CO). Leachate samples after appropriate

dilution were digested in a COD reactor for 2 hours at 150 °C. Absorbance was read

35

with a spectrophotometer (Milton Roy Spectronic 20D) at 620nm wavelength and

concentration was determined using standrads.

The pH of leachate samples was measured in the ﬁeld using a pH probe (Cole Parmer,
IL) and pH meter (Hanna instruments, RI) after calibrating the pH meter using buffer
solutions of pH 4, 7, and 10. Eletrical conductivity (BC) was measured using YSI
conductivity meter (30/25 FT). “Check standard” (Hanna Instruments conductivity

calibration solution 12.880 mS) was used to standardize the meter.

VOC was measured using gas chromatography mass spectrometer (GC-MS). The GC
used was (Agilent 6890) with Agilent 5973 network mass selective detector and J&W

Scientiﬁc 30mX0.250mm column (DB-624 Liquid phase coating). Column temperature

was held at 40 °C for 8 minutes, then increased at 5°C/min to a ﬁnal temperature of 200

°C. The method was same as EPA prescribed method for measuring VOCs.

3.4. Experimental Procedures

To accomplish the four objectives of this research, a series of experiments were
conducted on Northern Oaks bioreactor landﬁll. Figure 3-8 shows the title and duration

of each experiment followed by the associated sampling events.

36

35>“... wow—mama 98 3225“ 358598 mo 32225 “arm «.5»:—

 

 

 

 

 

 

 

Gm»

 

 

 

 

 

 

 

 

88

 

a:

 

 

 

 

 

 

 

963288 302550

93695
95% Sea mew—anew 8383

a: $55 mam—9:8 03:83
6% 35.6 mam—95m mew

nouoo E be Bow mmmcowoqaﬁowz
v.68 cowoﬁsmaoo commxo

N “moo cocoa E has

_ $2 compo E .__<

95.8% waist mmmoeowocmﬁoE

oomosemeoo E23

 

 

oo:

ooo _

ooo oow

ooh

ooo

oom

oov

oom

ooN

oo_

37

3.4.1. Performance of intermittent air injection system (Objective 1)

Objective 1 was to evaluate the intermittent air injection in terms of its impact on
temperature, oxygen and nitrogen dissemination into the landﬁll, and leachate
characteristics. Air injection was accomplished using a Fugi Ring compressor with
capacity 2.5 hp (500p-2T) with a maximum pressure of 80 inch and maximum ﬂow of
154 scfrn at two different time points. Air injection test 1 was conducted from Sep 03 to
Oct 03, while air injection test 2 was conducted from May 04 to Dec 04. Air injection test

1 occurred before the operation of the gas extraction system.

Hence, during air injection test 1, the horizontal gas extraction lines were used for air
injection and the horizontal leachate collection lines were used to vent the injected air.
Air injection test 1 had two phases. Air injection during phase 1 was conducted through
pipes in lift 1, while during phase 2 air was injected through lift 4. Air injection test 2
was conducted after the installation of the gas extraction lines. Hence leachate
recirculation lines were used to inject air. Air injection test 2 had two phases as well.
During phase 1 air was injected through lift 1 and during phase 2 through lifts 3 and 4.
Detailed description of each test is presented below. During the two air injection tests,
gas samples collected from most of the ports were analyzed for oxygen, nitrogen, carbon
dioxide and methane. In addition leachate samples were collected before and after air
injection experiments to evaluate the effect of air injection on leachate quality. Leachate
samples were analyzed for COD, pH, and VOC. Composition of the vented gas and the
ﬂow rate of injected air were measured. Ambient and landﬁll temperature and humidity

were monitored continuously.

38

Air injection test 1 was conducted between 9/5/03 and 10/24/03. During Phase 1, air
was injected into gas extraction line G12 and vented from leachate recirculation lines L32
and L42. The air was injected at a rate of 91.5 scfrn for 21 days, from 9/5/03 to 10/2/03.
The volume of air injected during Phase I was ~2.59 million cubic feet. During Phase 11,
air was injected into gas extraction line G4 and vented from leachate recirculation lines
L32, L5, and L6. The air was injected at a rate of 82.7 scfm for 22 days, from 10/2/03 to

10/24/03. The volume of air injected during Phase II was also ~2.60 million cubic feet.

Air Injection test 2 was conducted between (5/25/04 to 12/8/2004). During Phase 1 of
air injection test 1 (5/25/04 and 10/14/04), lifts L11, L12, L13, and L14 were used for air
injection. The active aeration period lasted for 130 days and air injection ﬂow rates
varied for each line ranging from 41 to 133 scfrn. The volume of air injected during this
period was 12 million cubic feet. The percentages of air-injected volume to each line in
comparison to the cumulative volume was 36, 26, 16, and 22 % for lines L11, L12, L13,
and L14, respectively (Table 3-2). This was calculated based on ﬂow rate in each line.
Phase 2 of air injection test 2 spanned from 10/15/04 and 12/08/04 and targeted lines
L31, L32, L33, L41 and L42 in lifts 3 and 4. The active aeration period lasted for 45
days. Air injection ﬂow rates varied for each line ranging from 46 to 122 scfrn. The
volume of air injected during this period was 5.77 million cubic feet. The percentages of
air volume injected in each line as a percent of total air injected was 21, 17, 8, 9 and 45%
for lines L31, L32, L33, L41, and L42, respectively (Table 3-3). During air injection test

2 (both phases), the gas extraction line was operational.

39

Table 3-2: Air injection test 2 (lower lift) - Percent ﬂow in each injection line

 

Distribution of air

 

 

 

 

 

 

. . Volume Duration of aeration
Injection pipe injection volume
(scfm) (Day)
(%)
L11 4,331,440 36 39
L12 3,072,482 26 39
L13 1,962,510 16 20
L14 2,621,045 22 32
Total 1 1,987,476 100 130

 

Table 3-3: Air injection test 2 (upper lifts) —injected air volume distribution among the

injection lines

 

 

 

 

 

 

 

 

Distribution of air Duration of
hip???“ Vr()lume injection volume aeration
“0 (%) (Day)

L31 1,224,891 21 6.9

L32 981,675 17

L33 459,284 8 4

L41 519,737 9 7.8

L42 2,588,227 45 19.4
Total 5,773,815 100 44.7

 

 

Computation of heat generation factor amounts to performing a heat balance around the

bioreactor landﬁll. The data for this objective comes from the two air injection tests.

Parameters needed to calculate the heat generation factor include, temperature and

moisture of the waste, temperature beneath the landﬁll, the ambient air temperature and

humidity, the air injection ﬂow rate, the gas venting composition, and the volume of the

40

 

waste. The approach to compute the heat generation factor is described in more detail in

section 3-6.

3.4.2. Oxygen consumption rate (Objective 2)

The experiment to calculate oxygen consumption rate was conducted during the air
injection test 1 in 2003. Brieﬂy, after maintaining air injection for several days, it was
stopped so that decrease in oxygen concentration at each sampling location could be
monitored. A background sample was taken just before stopping the air injection, and
then gas sampling continued until the oxygen levels at the sampling locations dropped to
very low concentrations. This experiment was repeated twice at different initial
temperatures, moisture contents and starting oxygen concentrations using sampling ports
located in lifts 1, 2, 3 and 5. Lifts 1, 2 and 3 were at similar moisture content as given by

moisture probes however lift 5 had lower moisture content compared to all others.

3.4.3. Evaluation of anaerobic digestion process after startup or air perturbation
(Objective 3)

Anaerobic digestion process was evaluated at the start of the bioreactor landﬁll cell
operation (i.e., prior to air injection test 1) as well as following air injection test 1. It was
accomplished by measuring methane and hydrogen concentration in the sampling ports
inside the bioreactor landﬁll. During the startup phase of the bioreactor landﬁll
(1/19/03- 9/5/03), 19 gas sampling tests occurred. One sampling event amounted to
collecting samples from most of the 48 locations for gas and leachate analysis. Gas and
leachate monitoring after air injection lasted up to 2 months. Temperature within the

landﬁll was continuously measured during this period. Samples were analyzed for

41

methane, carbon dioxide, oxygen, nitrogen and hydrogen by methods described above.

Leachate samples were analyzed for VOC, COD, and pH.

3.5. Statistical Methods

Statistical analyses were conducted using SPSS version 11. The tests used to analyze the
data include: t-test, Mann-Whitney U Test, paired t-test, Wilcoxon T Test, Pearson and
Spearrnan's correlation, and partial correlation. T-test and Mann-Whitney U Test were
used to test the difference in the mean and median of two independent variables,
respectively. Paired t-test and Wilcoxon T Test were used to compare two dependent
variables in terms of their mean and median, respectively. Pearson and Spearrnan’s
correlations study the signiﬁcance and strength of a linear relationship between two

variables.

The above tests are grouped into two types; parametric and nonparametric (Table 3-4).
Parametric tests are conducted when the normality assumption is veriﬁed. The normality
assumption means that the data is normally distributed. For non-normal distribution,
parameters of estimation used to describe the normal distribution such as mean and
standard deviation cannot describe the data any more. Therefore, nonparametric tests do

not use these parameters.

All nonparametric tests require ranking of the values. In order to do so the values from
both samples are combined and ordered from lowest to highest. A rank is given for each

value. In case of a tie in the values of the observation, the range of the ranks for the

42

group will be calculated, and an average rank for each observation in the group will be

 

 

 

 

given.

Table 3-4: Comparison between parametric and nonparametric tests

Reason to test Parametric test Non-parametric
analogue
[1; est difference Independent Mann-Whitney U
etween two
. samples t-test Test
independent samples
TCSt difference Paired sam les t-
between two dependent test p Wilcoxon T Test
samples
Tests the linearity of a , ,
. . Pearson s Spearrnan 5

relationship between . .

. correlatlon correlation
two varlables

 

 

 

3.5.1. Normality assumption

Several methods are available to test the normality assumption. The ﬁrst method
involves calculation of skewness (3-1) and kurtosis (3-2) of the variable and visual
observation of its distribution. Kurtosis is the measure of whether the data are peaked or
ﬂat in comparison to a normal distribution. Skewness is a measure of lack of symmetry.
If the range of kurtosis and skewness is between -1 and +1, then the assumption is valid.
If these parameters are not within the normality range, the central limit theorem can be
applied. This theorem indicates that if the number data points is above 50, then

parametric tests can be used.

_ 4
2:100 " y)
(n —1)SD4

 

kurtosis = — 3 ................................................................................... (3-1)

43

where,

E is the mean,

SD is the standard deviation, and

n is the number of data points.

n -3
Zi=](yi -y)

3 ....................................................................................... (3-2)
(n —1)SD

 

skewness =

The other two methods to test the normality of distribution of the data are Kolmogorov-

Smimov (D) given by equation (3-3) and Shapiro-Wilk tests (W) given by equation (3-

4).

D = max (F(Y,-))— i ................................................................................................ (3-3)
lSiSn n

where,

F is the theoretical cumulative normal distribution

’1
Zaixo)
'=l

W = —n’————— ...................................................................................................... (3-4)
(x,- " 3‘22
E

where,

x(i) are the ordered sample values (x(l) is the smallest), and

a,- are constants generated from the means, variances and covariances of the order

statistics of a sample of size n from a normal distribution.

3.5.2. Test difference between two independent samples

The two tests used to evaluate the difference between two independent samples are t-test
and Mann-Whitney U Test. The t-test is used when the distribution of data in the two
samples is normal; otherwise Mann-Whitney U Test is used. The t-test is conducted on
the sample values directly, while Mann-Whitney U Test is conducted on ranked data. In
order to conduct t-test, t given by equation (3—5) and the degrees of freedom (df) are
calculated. Then using values of (t) and (df), the signiﬁcance of the test is read from
statistical tables. If the signiﬁcance of the test is less than 0.05, then there is a signiﬁcant

difference between the two samples being tested. The standard error of the difference in

mean (SE diﬁ" ) used in equation (3-5) and the degrees of freedom are calculated by either

of the following methods. The choice of method depends on the homogeneity of the
variances of the two samples. Homogeneous variances mean equal variance in both
samples that are tested for difference in their mean. These equations for the two cases are

shown in Table 3-5.

 

M .

t= “ff .................................................................................................................. (3-5)
SEdl-ﬂ‘

where,

Mdiﬂ =|M1—M2 ................................................................................................... (3-6)

45

M1 and M 2 are the means of ﬁrst and the second samples, respectively.

SE at)? is the standard error of the difference in mean (Table 3-5)

Table 3-5: Calculation of the standard error of the difference and the degree of freedom
for the t-test under homogeneous and non-homogeneous variances

 

. Non-homogeneous
Homogenous variances

 

 

 

 

 

 

 

 

 

 

variances
Standard error of the 2 2
SD SD
_ l 2
difference SE diff = (SDI + SD2 ) SEd'ﬂ —\/ n + n
J n1 + n2 1 2
2 2 2
[SD/ 313/]
+
”1 "2
Degree of freedom _ _ _ df = 2 2
df - (n1 1) + (m 1) 5012 3022
"1 "2
+
n1 — n n2 —1
Were,

n1 and n2 are the sample size of ﬁrst and the second samples, respectively, and

SDI and SDZ are standard deviations of the ﬁrst and second samples

In order to conduct Mann-Whitney U Test, U value is calculated (3-7). If U value is
bigger than that in the statistical tables that corresponds to the number of the sample
points at the 0.05 signiﬁcance, the test is signiﬁcant.

M-R, ......................................................................................... ,(3-7)

U=nn +
12 2

46

 

where,

R] is the sum of the ranks in sample 1, and
n1, n2 are the sample size of 1St and 2nd samples.

3.5.3. Test difference between two dependent samples

The difference between two dependent samples can be evaluated by using either paired t-
test or Wilcoxon t-test. Paired t-test is used when both samples are normally distributed;

otherwise its nonparametric analogue (Wilcoxon t-test) is used.

To conduct a paired t-test, the degrees of freedom (df) given by equation (3-8) and t value
given by equation (3-9) are calculated. Using statistical tables, the signiﬁcance that
computed t value and the degrees of ﬁ‘eedom is read. If the signiﬁcance is less than 0.05

then there is a signiﬁcant difference between the two groups of samples.

df = n —1 ................................................................................................................... (3-8)

Where,

n is the number of the pairs.

 

t = d .................................................................................................................. (3-9)
SE diﬂ
Where,

47

E is the mean of the difference in value of the paired data points, given by equation (3-

10), and

SEdiﬁ‘ is the error of the difference, given by equation (3-11)

3:21—49 ................................................................................................................... (340)

n

Where,

d,- is the difference in value for the paired data points

 

SEW = ........................................................................................................ (3-11)

SDdiff is the standard deviation of the difference in value of the paired data points, given

by equation (3-12)

._— 2
2(‘1' d) ............................................................................................. (3-12)

n—l

SDdIﬂr =

To conduct Wilcoxon t Test the data is ranked as discussed above for all nonparametric

tests. If there are no ties in the ranking, then equations (3-13) through (3-15) are used.

2 J31?) .............................................................................................................. (3-13)

J VAR

48

Where,

VAR is the variance of the expected rank, given by equation (3-14)
ER is the expected rank sum, given by equation (3-15)

R is the higher rank sum of the two samples, given by equation (3-16),

 

VAR = "(n + ”(2" +1) ............................................................................................... (3-14)
24
ER = w ............................................................................................................ (3-15)

There may be two kinds of ties in this test. In the ﬁrst case the values of the pair are
exactly the same. In the second case, the difference in the values of the pair is the same.
In the ﬁrst case, the measured values are ignored and n is adjusted accordingly. In the
second case, the variance calculated by equation (3-14) will be reduced by a value of x,

which is calculated by equation (3-16).

 

3
t -t
= ................................................................................................................... 3-16
x 48 ( )
Where,

t is number of the tied ranks.

3.5.4. Tests the linearity of a relationship between two variables

Quantifying the linearity of relationship between two variables is established by Peasron

or Spearman’s correlations. When the dependent variable is normally distributed,

49

 

Pe'

Cl

Pearson correlation (ryx) is used given by equation (3-17). Otherwise the nonparametric
analogue, Spearman’s correlation (r,) is used, given by equation (3-19). To conduct
Spearman’s correlation values of samples are ranked, and the Spearman’s correlation is

calculated based on the ranks.

EPG-50’1"?) S
r ="=‘ = "y ...................................................................... (3-17)

y" Jams}, 5,0,3”

Where

 

Sxx, Syy are the variances of x and y variables, respectively. Variance is given by

equation (3-18),

3 and E are the average of the values of the variables y and x, respectively, and

n is the number of samples.

n —
SW = Z(y,~ -y)2 .................................................................................................... (3-18)
i=1
6 d2
,._..-z_ ....................................................................................................... .3-...
n(n -1)
Where,

d is the difference in ranks between the two variables

50

Partial correlations quantify the linear correlation between two variables while

controlling for a third one. The formula to determine the partial correlation coefﬁcient

ryxc, where c is the control variable, and y is dependent on x is given by equation (3-20).

If partial correlation equals the correlation without control of variable c, then the control
variable does not affect the correlation of variables x and y. A partial correlation that
equals zero indicates a false correlation between variables x and y, and the relationship is

not direct between these variables.

ryx—rycrxc

= ............................................................................................ 3-20
ryxc ‘ll _ ryc J1 _ rxc ( )

 

 

3.6. Method to Determine Heat Generation Factor

Heat generation factor is deﬁned as the amount of heat produced (kJ) per unit mass of
oxygen (kg) consumed. To compute heat generation factor, a heat balance around the
bioreactor landﬁll needs to be performed. Heat balance equation states that
accumulation of heat equals to the total heat gained or produced minus the total heat loss.
Sources of heat generation in a bioreactor landﬁll are microbial exothermic reactions and
sensible heat gain. Losses in heat can be attributed to sensible heat, latent heat and heat
loss to boundaries. When including these heat losses and gains the heat balance equation
would be accumulation of heat = i sensible heat — latent heat + metabolic heat generation
- heat loss to boundaries

Mathematically the above equation can be written as:

q accumulation = 313‘] sensible i qlatent + q metabolic i q boundaries ---------------------------------- (3'21)

51

The following assumptions are made to solve equation (3-21):

The solid and gas phases are in equilibrium. This means that the temperature of
air and waste are equal in the bioreactor. This is usually named as homogenous
assumption, and is frequently used when conducting heat modeling in composting
operations[24].

The gas phase within the bioreactor landﬁll and the efﬂuent gas are saturated with
water vapor. This is true because at low aeration rates and high waste moisture
content, moisture transfer from solid waste matrix to air encounters no resistance
[24].

Changes in the pore space and reactor working volume were assumed to be
negligible during the period of observation used in the heat balance study. This
assumption is justiﬁed because the duration of each air injection is less than 23
day.

Oxygen enters the reactor only through forced aeration (advection ﬂow) and the
diffusion of oxygen into the landﬁll is negligible. The landﬁll cell was capped
with HDPE geomembrane minimizing oxygen diffusion.

Efﬂuent gases are at a constant pressure. This assumption also been used in
previous studies [24]. Pressure change in the landﬁll is measured and was in the
range of 1- 2 mbar approximately about 0.2% of the atmospheric pressure.

The moisture content of the solid waste matrix is assumed to be constant.

The heat generation factor is constant and is a result of aerobic reaction only, and

heat generation due to anaerobic process is negligible [74]. Although some

52

anaerobic reaction might be taking place, the heat produced of aerobic reaction is

much higher than of the anaerobic one.

Using the above assumptions and substituting with the appropriate equations for each

heat term in equation (4-2), the heat balance equation is given by equation (3-22).

d
pCpVEU") =FLCprL(TL "Twaste) +FaiGCairpair(Tair ”Twaste)+

............ (3-22)
+ AHw(xw,out — xw,in ) + QR(02) + UA(Tout '— Twaste)
Each of the ﬁve terms is described in detail below.
1. Accumulation of heat is given by equation (3-23)
d
qaccumulan-on = pCszi-t-(T) ..................................................................................... (3-23)

where,
p is the apparent density of the waste (kg/m3),
V is volume of the waste in the landﬁll (m3),
T is temperature (°C),
t is time (sec),
dT/dt is the rate of temperature increase during air injection. This rate equals

the slope of temperature vs. time curve, and

Cp is the speciﬁc heat of solid waste (J/kg °C), given by equation (3-24)

Cp = Cp-solids °msolids + Cp—water 'mwater .............................................................. (3-24)

where,

53

Cp-so]ids is the heat capacity of solids and ranges between 0.5-2 J g'1 0C'1 for

solid waste [58]. A typical value of 1.2 J g.1 0C1 is generally used, including this
study,

Cp-wate, is the heat capacity of water 4.1818 J g.1 °C.1 [110],
mwatc, is the mass fraction of water, and

msonds is the mass fraction of solids.

2. Sensible heat change is the heat change in the landﬁll due to transport of ﬂuids and

air [55]. It is given the following equation:
qsensible = FL CpL pL (TL - Twaste ) + Fair Cpair pair (Tair '" Twaste) ----------------------- (3'25)
where,

Twaste is the temperature in the waste (°C),

Tai, is the temperature of the incoming air (°C),

TL is the temperature of the incoming leachate (°C),
F is ﬂow rate (m3/s),

Cp is the speciﬁc heat (J/kg°C),

p is density (kg/m3), and

subscripts air and L refer to the air or leachate moving injected in the

landﬁll.

54

3. Latent heat loss is due to evaporation of water, and can be signiﬁcant if humidity of

the injected air is low. It is given by equation (3-26)

qlatent = AHW(xw,om - xw,in) .................................................................................. (3’26)

Where,

AH“, (J/kg) is the heat evaporation of water,

x w, in is water mass in the coming air per unit time (kg/s), given by
equation (3-27), and
x w, out is water mass in the outgoing air per unit time (kg/s), given by

equation (3-28).

 

M P -
xw in = W W" Fair ..................................................................................... (3-27)
, Vstp P0 " Pw,in
MW PW,0W Tair

 

 

x , , = - ...................................................................... (3-28)
w ou Vstp PO _ Pw,0ut arr T waste

where,

MW is the molecular weight of water (0.01 8 kg/mole),

Vstp is the volume of one mole of water vapor at STD (22.4 L/mole),
P0 is the atmospheric pressure (atm),

Fair is the air ﬂow rate at STD (L/s)

Pw, out is the partial pressure of water inside the landﬁll and equals to saturated

vapor pressure at the waste temperature (atm),

55

T air is temperature of the incoming air (°C)

T waste is the temperature of the waste (°C)

Pw,in is the partial pressure of water in the incoming air (atm) and given by

equation (3-29),

RH 0
PW, in = 1—06 . w,in ...................................................................................................... (3'29)
where,

RH is the relative humidity in the incoming air (%), and

O

Pan is the saturated vapor pressure at the air temperature

4. Metabolic heat generation is given by the following equation:

qmetabolic = Q.R(02) ............................................................................................... (330)

Where,

Q is the heat generation factor (J /mol 0;), and

R(02) is the amount of oxygen consumed per unit time (mol Oz/s), given

by equation (3-31).

x - - X
Fair (%)
12(02) = V ............................................................................................. (3-31)
stp

 

Where,

x is the concentration of oxygen in the venting or extraction lines (%),

xai, is the concentration of oxygen in air (%),

56

Vstp is the volume of one mole of water vapor at STD (22.4 L/mole), and

Fair is air ﬂow rate at standard temperature and pressure (Us)

5. Heat loss to the boundaries is given by equation (3-32)

(1,0,, = UA(Tou, — Twm) ......................................................................................... (332)

where

U is the heat transfer coefﬁcient (J s.I m.2 -°C),
A is the area of the bottom of the landﬁll, footprint area (m2)
T waste is the waste temperature (°C), and

Tout is the temperature of soil undemeath the landﬁll (°C).

3.7. Method to Determine Oxygen Consumption Rate

Ideally, oxygen consumption rate should be determined in a manner that allows
normalization of the rate to the amount of organic matter in the waste. Under ﬁeld
conditions, assigning an active volume of waste to a given rate of consumption is not
feasible. Therefore, a new term, called “Point” oxygen consumption rate (POCR) was
coined in this study. Point oxygen consumption rate is deﬁned as the rate of decrease in
oxygen concentration with time at a given location in a bioreactor landﬁll. Although
POCR is not ideal, it has the advantage of measuring the consumption rate under ﬁeld
conditions. Neither oxygen consumption rate, nor POCR have been measured under ﬁeld

conditions.

57

The approach to determine POCR includes (1) oxygen concentration vs. time curve (data
collected), (2) calculation of liquid phase oxygen concentration, and (3) calculation of

kinetic coefﬁcients for POCR using Monod equation.

3.7.1. Oxygen concentration vs. time curve

The method to obtain POCR involved measuring oxygen concentration at a sampling
location for a period of approximately 8 hours. This period included a background
reading before stopping the air injection. Figure 3-9 shows an example of change in the
concentration of oxygen and other gas phase constituents concentration after stopping air
injection. The concentration of oxygen decreased from 13% to 2% in 7 hours. During
this time, carbon dioxide concentration increased from 4% to 9%. The nitrogen and

methane concentrations stayed constant during the entire period of the experiment.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

,HW . “r h—

2:: s
.o u:
:1

lg g
8 8

ii: 8

:0

e0 0

| 2‘

l

l o 1 2 3 4 5 6 7 8

E Time(hr)

I

|

:o— CH4 -—c02 +02 -x-N2

 

 

 

Figure 3-9: Gas concentration change after stopping air injection location 6143

58

3.7.2. Calculation of liquid phase oxygen concentration

POCR was computed using liquid phase oxygen concentrations. Gas phase oxygen
concentration was converted to liquid phase concentration and corrected for temperature
and salting out effect. For each sampling location, Henry’s constant was calculated based
on the corresponding temperature at that location, and corrected for the salting out effect.
Then Henry’s constant combined with the partial pressure of the oxygen at each location
were used to calculate the liquid phase oxygen concentration. The underlying assumption
when using this equation is that the gas and liquid phase oxygen concentrations are at

equilibrium. The liquid phase concentration of oxygen was calculated by equation (3-

 

33).
0 = y'P ............................................................................................................ (3-33)
100H2
Where,

y is the percent of oxygen in the gas phase at a given location (%),

H2 is corrected Henry’s constant for temperature and salting out effect (L atm/mole), and

P is the pressure within the landﬁll (atm). Pressure measured in the landﬁll at sampling
locations during the POCR study did not vary signiﬁcantly compared to atmospheric.

Hence, atmospheric pressure was used in equation (3-33).

Temperature correction for Henry’s constant

The Van’t Hoff -Type equation was used to correct Henry’s constant for temperature,

and is given by equation (3-34).

59

H2 =H1 exp-£[i—i] .............................................................................. (3-34)
R T2 T1

Where,

H1 is Henry’s constant at known temperature T1 (K),

H2 is Henry’s constant at unknown temperature T2 (K),

AH° is the standard enthalpy change in water due to the dissolution of oxygen in water

(kcal/kmol),
R is the universal gas constant (1.987 kcal/kmol.°K),

Another form of equation (3-34) is given by equation (3-35).

O

10307) = -

 

+ C ................................................................................................ 3-35
RT ( )

The constants C and AH° for oxygen in equation (3-35) equal to 7.11(tmit less) and 1.45

103 (kcal/kmol), respectively [111]. Using these values Henry’s constant (H) for oxygen

was calculated in units of atmosphere.

Henry’s constant units of (atm) can be converted to units of (L atm mole-1) by equation

 

 

(3-36).

H[ My") = ”(atm) 1 ................................................................................... (3-36)
m0 8 m0 3
MVwater( L j

60

Where,

M Vwam is the molar volume of water and equals to 55.6 mole L'l.

Salting out effect correction for Henry’s constant

As described in [112], correcting for the salting out effect on Henry’s constant is given by

equation (3-37).
H 2 = H17 .................................................................................................................. (3-37)

Where,

H2 is Henry’s constant corrected for both salting out effect and temperature (L atm/mole),

H1 is the Henry’s constant corrected for temperature in (L atm/mole), and

y is the activity coefﬁcient for oxygen solubility in salt water, given by equation (3-38).
log(7) = ks! .............................................................................................................. (3-38)

Where,

k8 is the salting out coefﬁcient and equals 0.132 for oxygen, and

I is the ionic strength, given by equation (3-39).

1 = 0.016EC .............................................................................................................. (3-39)

Where,

EC is the electronic conductivity [ﬁg].
cm

61

3.7.3. Calculation of kinetic coefficients for POCR using Monod equation

Monod kinetics, given by equation (3-40) is the most common model used for aerobic

substrate degradation [1 13-1 15].

V _ dS = _kmax.S.X

— — ............................................................................................... (3-40)
dt K S + S

Where,

V is the substrate utilization rate (mg L.1 h.1 ),
S is the concentration of substrate (mg L'l),

. . —l . . .
Ks rs the half saturation constant (mg L ); re the substrate concentration at which the

substrate utilization rate is half the maximum substrate utilization,

X is concentration of biomass (mg /L),

kmax is the maximum speciﬁc substrate utilization rate (L mg '1 h'l), and

t is time (h).

The kinetic coefﬁcients (kmax, K5) are classiﬁed into two groups based on the ratio of the

substrate to the biomass concentration. When the ratio of substrate to biomass is high,
unrestricted growth occurs and the kinetic coefﬁcients are called intrinsic. When the
ratio is small the kinetic coefﬁcients are called extant [116]. While determining extant

kinetics the growth is minimized thus the concentration of biomass X is relatively

62

constant [116]. When change in the biomass concentration is negligible, Monod equation

(3-40) can be reduced to equation (3-41) [117].

dS k.S
V = — = .................................................................................................... 3-41
dt Ks +S ( )

 

k is the maximum substrate consumption rate (h'l) , given by equation (3-42).
k = kmaxX ................................................................................................................ (3-42)

In order to obtain a solution for Monod equation (3-41), kinetic parameters (k, K.) have

to be unique and independent. Knights et. al. [118] found that a unique solution for

Monod equation can be obtained when the ratio of the initial substrate concentration (So)

/
to the half saturation constant (Ks) to be larger than 0.1 iii—0 > 0.1]. Moreover, the
S

uniqueness of estimates for k and Ks depends on the number of observations and the

precision of the measurements [118].

Estimation of Monod coefﬁcients k and KS is rather difﬁcult because these two kinetic

coefﬁcients are not completely independent [114]. The experimental conditions at

which these parameters are estimated can be controlled to lead to minimal dependence

among the parameters [114]. When the ratio of SOIKs is bigger than 1, a good separation

between the coefﬁcients k and Ks can be obtained [118].

63

To estimate the kinetic coefﬁcients, substrate concentration (S) must be distributed over a
large concentration range of the rate curve. This is ensured if the substrates

concentrations used cover the following three regions [114]:

1) Region in which the slope is very steep and constant (0 < S/Ks S2, 0 <V/k $0.67). In

this region ﬁrst order kinetics apply.

2) Region in which the slope changes rapidly (2 < S/Ks 39, 0.67 < V/k 50.9)

3) Region in which the slope is relatively ﬂat and near maximum utilization rate (S/KS

29, Wk 2 0.90). In this region zero order kinetics apply.

The two extreme regions under which Monod equation converts to either ﬁrst order or

zero order kinetics are shown below.

In the ﬁrst region, the concentration of substrate is low, i.e. (S << K s) then the

degradation rate will follow ﬁrst order kinetics with respect to substrate. This is given by

equation (3-43).

V=£=—£S-=—le ............................................................................................ (3-43)
dt KS
Where,

k1 — First order rate constant of substrate degradation (h")

64

The integrated form of equation (3-43) is:
S = S0 exp(—k1t) ....................................................................................................... (3-44)

In the third region, the concentration of substrate is high, i.e. (S >> K S) then the

degradation will follow zero order kinetics with respect to substrate. This is given by

equation (3-45).

dS kS
V=—=-—=—k=-k ........................................................................................ 3-45
dt S o ( )

Integrated form of equation (3-45) is:
S = S0 — kot ............................................................................................................. (3-46)

Where,

k0 — Zero order rate constant of substrate degradation (mg L'1 h'l)

Three methods to solve Monod equation (5-9) are: the integrated solution of Monod
equation, linearized form of Monod equation, and nonlinear regression methods. The
following two sections describe 1) the methods to solve Monod equation, and 2) ﬁtting

POCR data to the Monod equation.

Methods to solve Monod Equation

The ﬁrst method to solve Monod equation (3-41) is by its integration and then solving the
analytical solution by least square method [119]. Integration of equation (3-41) is given

in equations (3-42) through (3-45).

65

KS+S

 

 

 

— d5 = dt ...................................................................................................... (3-42)
'IESS—dS +%dS = —dt ................................................................................................ (3-43)
Kks 1n(SO)+,1C—SO—Kksln(S)-—}c—S=t .................................................... (3-44)
%1n(%9) +£(sO — S) =z ...................................................................................... (3—45)

The integrated solution given by equation (3-45) is in the form of time (t) as a ﬁmction of
substrate concentration. The least square method minimizes the difference between the

measured time and the calculated time by equation (3-45) for each measured substrate

concentration, using variable kinetic coefﬁcients k and Ks. In order to estimate the

coefﬁcients k and Ks by varying time rather than substrate concentration, measurements

of substrate concentration must be accurate.

The second method to solve Monod equation uses the linearized form of equation (3-41),
also known as Lineweaver-Burk plot [113, 120]. The linarized form of Monod equation

is given by equation (3-46).

-53.
k

+ ........................................................................................................ (3-46)

Cal—-
Pelv—

_1_
V

In this solution the rate dS/dt is substituted with a measured value of the substrate
utilization rate (V). To accomplish this, laboratory experiments involving concentration

vs. time curves with different initial substrate concentrations are conducted. Using these

66

curves, initial substrate utilization rates are calculated. These rates equal the slope of the
initial linear portion of the concentration vs. time curves (Figure 3-10). For each initial

substrate concentration, an initial rate is obtained. For example the rate in Figure 3-10
equals 3.23 mg L.1 h'1 at an initial concentration of S=8mg/L. (The data in this ﬁgure are
hypothetical and for the purpose of explaining the method). Substituting the rate (V) and
concentration (S) in equation (5-15) will result in a linear plot in which Ks/k is the slope

and l/k is intercept.

The advantage of this method is its mathematical simplicity. However, this method
suffers from two disadvantages. First it assumes that the errors are normally distributed,
which might not be valid. Second, this method cannot be applied under ﬁeld conditions.
This method has been used by many researchers in the past because of the complexity of

nonlinear regression methods that involve iterative techniques to obtain optimal values of
estimates of k and Ks [118]. With the improvement in the computerization techniques,

nonlinear regression techniques are now preferred.

67

 

 

    

 

 

 

 

 

 

9 .
I A 8 ' ——— —. _ .- ___, 2w, 7%
E}, 7 ,__l=-3.23x+7.8027_ , g
V 2—
g 6 ”.2 R;o.asa5__-_ -ﬂﬁ __
.g ____ __ __
g 5
G 4 ‘ .2_ _ _ _ _____ ___ ‘2-
§ § 1 ° l
l 8 3 p --+.—---— ——~ ﬂ ————~ ~——- *
2 w + h m m a
i ? 1 +__ '_'_—4"§".—; ' _——___ 7 "ﬂ — I
i O 0 l .——~ ————————?~ ’ -»—-+—--—0— o ~~ '1 i
‘ o 2 4 6 8 10 12 ,
i Trme (hr) E

 

Figure 3-10: A hypothetical example for determination of the initial rate for Lineweaver-
Burk plot

The third method to solve Monod equation (3-41) is by nonlinear regression analysis,
which is now routine due to improvements in computational methods [114, 118, 121-
123]. In addition, the use of this method does not require constructing several curves of
substrate vs. time. The rate can be determined using one curve. This is particularly
important when data collection is expensive or it is not feasible to obtain several curves.
In this method, given initial values for k and Ks the residual sum squares between the
observed and predicted concentration is minimized. The values of k and Ks when the

residual sum squares reaches minimum is the optimum solution.

In this study, the method chosen to solve Monod equation (3-41) was nonlinear
regression. The differential equation was solved numerically using Runge-Kutta

approximation technique. Given an initial concentration of oxygen and trial set of k and

68

Ks, the residual sum squares between the observed and predicted concentration was

evaluated using frnincon function in Matlab, given by equation (3-42).

 

 

’.’ s,—s 2 k .
f(6)= \[ZIJ ) ,9: K, ........................................................................ (3-42)
n so

Equation (3-42) attempts to ﬁnd a constrained minimum of a scalar function of several
variables starting at an initial estimate. The least square minimization routine based on
Levenberg Marquardt algorithm was used to minimize frnincon function. Coefﬁcients
reﬁning continued until frnincon reached global minimum. The coefﬁcients

corresponding to the global minimal value were taken as the best ﬁt.

69

4. RESULTS AND DISCUSSION

Low ambient temperatures in cold climates during waste ﬁlling period can lead to
persistent low temperatures inside a bioreactor cell. Air injection is a practical option to
increase waste temperature. In Northern Oaks bioreactor landﬁll, low temperature in the
waste was an issue leading to a slow start of the anaerobic process. The three objectives

in this study were:

Objective 1: To establish and quantify relationship between amount of heat produced
(1d) per unit mass of oxygen supplied (kg), and to characterize the performance of air

injection with respect to temperature and leachate characteristics.

Objective 2: To establish the spatial variability in point oxygen consumption rate
(decrease in oxygen concentration with time at a given location in a bioreactor landﬁll),
and study the effect of temperature, oxygen concentration and moisture content on point

oxygen consumption rate.

Objective 3: To quantify the performance of anaerobic digestion at startup and following
air injection with respect to methanogenesis establishment, using methane and hydrogen

concentrations as indicators for anaerobic digestion process.

Two air injection tests for evaluation the correlation between air injection and
temperature increase were conducted. POCR was determined during the ﬁrst test by

stopping the air injection for about 8 hours. Methanogenic activity was studied before air

70

injection test 1 (i.e., during startup of the bioreactor landﬁll) and after the air injection

test 1 was stopped.

4.1. Evaluation of Intermittent Air Injection in Northern Oaks Bioreactor Landfill

Two separate air injection experiments were performed to quantify relationship between
increase in waste temperature inside and amount of air injected. This section starts by
establishing the base line for waste temperature before air injection. Then evaluation the
response of temperature to air injection is described in the second subsection. The third
subsection addresses the response of waste temperature response after stopping air
injection. In the forth heat generation analysis are presented. The ﬁfth section describes
the gas phase concentration—time proﬁle during the air injection. In the last section, the

impact of air injection on leachate characteristics is characterized.

4.1.1. Establishment of base line temperature before air injection

Base line of temperature was established by studying the ambient temperature and its
relation to initial waste temperature. In addition, the effect of leachate recirculation

before air injection on waste temperature is discussed in this section.

Correlation between waste and ambient temperature

Low temperatures such as those observed in Northern Oaks bioreactor landﬁll have not

been reported in literature before. Temperatures reported in landﬁlls in the range
between 10 and 70 °C [15, 17, 33, 34, 36-39]. Lower ambient temperatures during

ﬁlling and before waste coverage resulted in lower initial waste temperatures (Figure 4-

1). In this study, the average of waste temperature until ﬁve days after covering it with

71

subsequent waste is termed initial waste temperature. Figure 4-1 illustrates the average

of initial waste and ambient temperature for the ﬁve lifts.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1'- 95% conﬁdence interval of mean 0

Mean

 

:3 I: .. ”1337-7:
e: 20 he  ———~*~ my.
5 15: - / -—
§ 10 I "T ‘Tm‘Liﬂ3
g 5 i 'Liﬂ4wi7(W—_ny=l.6n301x+10.858
g 5 i- / _ _ RITOL9855

-15 -10 -5 0 5 10 15 20
Airtemperature(C)

 

Figure 4-1: Average waste temperature vs. Average ambient temperature waste is

exposed to in each lift

The range and average of ambient temperature for which waste was exposed to were (0.5

to 32 °C, 18:0.3 °C), (-8 °C to 20 °C, 38:03 °C), (-24 °c to 10 °C, 2810.4 °C), (-28

°C to 10 °C, -6.5i0.4 °C), and (-30 °C to -1 °C, -103-_0.4 °C), for lifts 1, 2, 3, 4 and 5,

respectively. The difference in ambient temperatures that each lift was exposed to was
signiﬁcant as determined by Kruskal-Wallis Test (Table 4-1). The resulting initial waste
temperatures reﬂected the above variation in ambient temperatures. Waste temperatures

were signiﬁcantly different among the lifts as determined by Kruskal-Wallis Test (Table

72

 

4-1). Average initial waste temperature in lifts 1, 2 3, 4, and 5 were: 27.73205 oC,

17.0:10 °C, 54:27 °C, 20:13 °C, and -6.5i1.3 °C, respectively (Figure 4-2).

Table 4-1: Kruskal-Wallis Test for testing differences between lifts waste lifts with
respect to a) ambient temperatures to which waste was exposed and b) waste temperature

 

 

 

 

 

 

 

 

 

 

Lift N Mean Cbi- d f Asymp.
number Rank Square Sig.
Lift 2 816 2625
- Lift 3 911 1750
A" 1543 3 <0.001

temperame Lift 4 1055 1211
Lift 5 502 758

Lift 2 69 103

Lift 3 25 56

waste 103 3 <0.001

temperature Lift 4 30 40
Lift 5 16 9

 

 

73

50

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

4o.
8 30‘
E 1—
g 20: I at
8" —— “l 6
O
E 101 —-[—
.93. I
53 0. ﬂ ‘ i
3 _r_
._.__-______
-101
-20 . -
N = 375 69 25 30 16

Liftl Lift2 Lift3 Lift4 Lift5

Figure 4-2: Initial waste temperature in the landﬁll

Initial waste temperatures were higher than ambient temperatures to which waste was

exposed. The differences between ambient and waste temperatures were: 9.6, 13.2, 8.2,
8.4, and 3.6 °C for waste lifts 1 through 5, respectively. Waste temperature correlated

strongly with ambient temperatures (R2=0.99). The relation between initial waste
temperature and ambient temperature to which waste was exposed for lifts 2, 3, 4 and 5

was established in this study as:

Twa,,,=1.6 Tambiem + 10.9 ...................................................................................... (4.1)

Where

Twaste is the waste temperature in a lift (°C), and

74

Tambiem is the ambient temperature the waste was exposed to in that liﬁ (°C).

Lift 1 was excluded from the analysis, because it was uncovered for a period of 90 days,
however lifts 2, 3, 4 and 5 were exposed to air for about 2 weeks each, making them

comparable.

The increase in waste temperature above ambient temperatures was attributed to heat
generating aerobic biodegradation reactions. Measured oxygen and carbon dioxide
concentrations in the landﬁll indicated that oxygen consumption and carbon dioxide
production were taking place in the waste mass. The oxygen and carbon dioxide
concentrations proﬁles in the gas phase for lift 4 are shown in Figure 4-3. Oxygen
concentration decreased from 9.0i2.9 % to 4.4i1 .7 % within two days after coverage and
further decreased to 1.7i0.3 % by day 18 after waste coverage. This decrease in oxygen
concentration was accompanied by an increase in carbon dioxide concentration from

20.7:70 % to 30.6i7.7 % during the ﬁrst two days, and to 33.71:] .67 % by day 18.

75

 

 

 

 

 

 

 

 

 

Carbon—dixoide concentration (%)i ’

 

 

 

 

 

j Days after covering lift 4

 

 

.. _.._ _1

E mm -rmx omedian —-—mean;

 

Figure 4-3: Gas composition after waste coverage

Effect of leachate recirculation on waste temperature

Leachate recirculation is used in bioreactor landﬁlls to enhance biodegradation. However
the effect of leachate recirculation on waste temperature is not yet documented. Before
the impact of air injection on waste temperature can be studied, it is important to
understand the effect of leachate injection on waste temperature. Leachate injection
started as soon as the waste was covered. The temperature of the injected leachate was

constant at about 125°C throughout the leachate recirculation period.

After the start of leachate injection, the difference between waste and leachate

temperatures began to decrease. When leachate was not practiced in a lift waste

76

temperature remained constant (Figure 4-4). Waste temperatures decreased in lifts 1 and
2 and increased in waste temperature in lifts 3, 4, and 5 as a result of leachate injection.

Average waste temperatures on the last day of leachate recirculation (9/5/03) were
10.9:06 °C, 12.5i2.4 °C, 10.1:42 °C, 10.4i4.1 °C, and 1.5:r4.0 °C, for lifts 1, 2, 3, 4,

and 5, respectively. The waste temperature in lifts 1, 2, 3, and 4 was not statistically
different after this period of leachate recirculation as determined by ANNOVA test
(Table 4-2). The volume of leachate injected expressed as a percent waste volume for

each lift were 8.1, 7.8, 7.4, 8.5, and 4.9 %, in lifts l, 2, 3, 4, and 5, respectively.

Table 4-2: AN OVA test for waste temperature differences for lifts 1 through 4 after
leachate recirculation (9/5/03)

 

 

 

 

Sum of Mean .
Squares df Square F 813'
Between 32.48 3 10.8
Groups
Wlthln 360.20 38 9.5 1.142 0.344
Groups
Total 392.68 41

 

 

 

77

noun—3:8» 8382 meta—u 8356.58“ he EH a: some 5 053.6988 8mm? owﬂo>< ”Viv ensur—

 

 

 

658688 2383 53862 2383 Em - 1.. Base 6883 2283 ll
_ 5' N a: .. m as .
:3 x :E . 98H?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(3) armatodura 1,

 

 

78

4.1.2. Effect of air injection on waste temperature

During test 1, the waste temperature in lift 5 increased from ~O °C to more than 30 °C. In
the remaining lifts 1, 2, 3 and 4 the temperature increased from 11.23209 °C to 17.8i2.3
°C (Figure 4-5). During air injection test 2, waste temperature increased from
14.5i1.2°C to 29.0:39 °C. The rate of temperature increase was calculated for a total
of 26 air injection events that took place as part of air injection tests 1 and 2. Detailed
description of these air injection events including their duration, start and end date, air
ﬂow rate and the location of the air injection line are summarized in Table 4-3. The rate
of waste temperature increase was calculated as the slope of the linear trend line ﬁtted to
temperature data over the duration of each air injection event. An example of this
calculation is shown in Figure 4-6 for two air injection events. The ﬁrst air injection
event shown in the ﬁgure was initiated on 7/2/2004 and continued for 16 days (air
injection event 9 in Table 4-3). Three days later, the air injection resumed again (air
injection event 10 in Table 4-3). The rate of temperature increase was calculated for the
two events separately. The rate of waste temperature increase for event 9 was 0.28
°C/day and for event 10 was 0.11 °C/day. The average and median of all rates of
temperature change in the bioreactor landﬁll during both air injection tests was 0.10:0.04
°C, and 0.05 °C/day. The average and median rates of temperature change were:
(0.14:0.08, 0.09), (0.10:1:0.08, 0.05), (0.14:0.08, 0.07), (0.05:t0.05, 0.03), and
(0.08:0.05, 0.04) °C/day, for lifts 1, 2, 3, 4 and 5, respectively. Table 4-4 displays the

26 air injection tests and the rate of temperature change calculated for each lift.

79

Table 4-3: Operational conditions for each air injection event

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Chronological Start End date Duration Injection Air ﬂow rate
numbering date (day) line (scfm)

Injection through lift 1

1 9/5/2003 9/7/2003 2 G12 87

2 9/8/2003 9/16/2003 8 612 89

3 9/22/2003 9/29/2003 7 612 94

4 9/30/2003 10/2/2003 2 G12 97

6 5/25/2004 6/17/2004 23 L11 41

7 6/17/2004 6/25/2004 8 L12 65

8 6/25/2004 7/2/2004 7 L13 70

9 7/2/2004 7/18/2004 16 L14 56

10 7/20/2004 8/5/2004 16 L14 58

11 8/5/2004 8/18/2004 13 L13 67

12 8/19/2004 8/24/2004 5 L12 61

13 8/24/2004 9/2/2004 9 L12 55

14 9/2/2004 9/3/2004 1 L1 1 127

15 9/7/2004 9/8/2004 1 L1 1 1 16

16 9/9/2004 9/17/2004 8 L11 124

17 9/20/2004 9/23/2004 3 L1 1 134

1 8 9/24/2004 9/27/2004 3 L1 1 130

19 9/27/2004 10/14/2004 17 L12 49
Injection through lift 3

20 10/15/2004 10/22/2004 7 L31 122

21 10/22/2004 10/24/2004 2 L32 91

22 10/26/2004 10/31/2004 5 1.32 98

23 10/31/2004 1 1/1/2004 1 L33 82

24 11/8/2004 1 1/11/2004 3 L33 88
Injection through lift 4

5 10/2/2003 10/24/2003 22 G41 83

25 11/11/2004 12/2/2004 19 L42 94

26 12/2/2004 12/8/2004 8 L41 46

 

80

 

Table 4-4: Rate of temperature change for air injection events in each lift

 

Rate of temperature change (°C/day)

 

Chronological

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. Lift 1 Lift 2 Lift 3 Lift 4 Lift 5
numbering
Injection through lift 1

1 0.13 0.13 0.00 0.01 0.00

2 0.24 0.21 0.15 0.00 0.00

3 0.44 0.58 0.10 0.04 0.01

4 0.37 0.60 0.21 0.02 0.02

6 0.18 0.22 0.15 0.00 0.01

7 0.07 0.12 0.03 0.00 0.03

8 0.04 0.40 0.00 -0.04 0.02

9 0.05 0.28 0.09 0.01 0.03

10 0.05 0.1 l 0.07 0.00 0.04

1 l 0.1 1 0.24 0.07 0.00 0.05

12 0.15 -0.11 0.01 0.05 0.04

13 0.14 -0.08 0.06 0.03 0.04

14 0.44 -0.14 0.00 0.00 0.03

15 0.47 -0.08 0.09 0.05 0.06

16 0.44 -0.03 0.08 0.02 0.05

17 0.47 -0.05 0.05 0.00 0.04

18 0.33 0.00 0.00 0.03 0.05

19 0.07 -0.02 0.13 0.03 0.05
Injection through lift 3

20 -0.10 0.04 0.41 0.03 0.00

21 -0.04 0.01 0.58 0.03 -0.03

22 -0.10 0.06 0.35 0.06 -0.23

23 -0.04 0.08 0.77 0.06 0.06

24 -0.10 0.11 0.07 0.05 0.06
Injection through lift 4

5 -0.06 -0. 10 0.07 0.60 1.56

25 -0.05 0.03 0.07 0.13 0.06

26 -0.05 0.04 0.00 0.15 0.07

 

81

 

Eggs 88385 830 805.52 5 ma: 2:.“ E ogﬁomﬁou 8m“? owmco>< “m4 95E."—

 

 

 

 

 

 

 

 

 

 

 

 

 

(Q) armamdural

m5.-le§t-lm§r 1.35-1-5: e88E8a2m1-|=88.§e< magi- M
$63 ea:
82 82 8» Se 84 com o
_ _ _ — 2 owl
i ON!
o
8
h 9.
_
i 25 u 8
m e: N a: _
All a HWDH.
Nash :ﬁ

 

 

 

 

 

cw

 

 

 

82

 

 

 

 

 

 

y=0.11x+ 26.27

 

*—_b—

 

 

 

 

 

 

 

 

 

 

 

I
i
6 l 2
V R = 0.98
i —*—+~we~~-
e I |
E 23 ,_______..___ —_——-— —— 1 1 e — ”_.__— _.__
3 I y = 0.28X + 24.42 I I
g - R2 = 0.99 J l
21 _g _L _.__ _Lw .___,__~ .2“ _ _ . _.___ _W M
3 ' | I
3’9 |
g 19i——— —- ———~ --' I — e-
3: i | I
| I
17 FT- .Wwv __ _._ .. . _ ‘
l l l
15 i——— r r 4.- r t i i
0 5 10 15 20 25 3O 35 40
Days from starting air injection
F.

[— — Airiry'ectionpaused o Event9 I Event 107‘

 

 

Figure 4-6: Temperature increase in waste during air injection in L14 for 1ift2

When air injection was conducted through lifts 3 and 4, waste temperature in the lower
lifts decreased or remained constant. The decrease in waste temperature was due to net
heat loss to the bottom of the landﬁll. When air injection was conducted through lift 1,
the temperature did not change signiﬁcantly in the upper lifts. During air injection test 1
air was initially injected through lift 1 and shifted to lift 4 after 27 days. The temperature
change that corresponded in all lifts during air injection test 1 is shown in Figure 4-7.

The start of air injection in lift 1 is shown by the vertical dotted line at day 420. Air

83

 

 

 

 

Temperature (C)

 

40

 

 

   
   

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

1 é-if:
o - a...» - 4
F i'. au 1
i 5 I ‘1 y i:
I I i'
-5 i j
. , l
'10 7 l I 1 l
400 420 440 460 480 500
Time (days)
------ —AirTemp 1' Lift5
x Liﬁ4 A Liﬁ3
I Lift2 —-Lift1
i —-—Stop air iry'ection — - - Start air injection
17 - - -Switchtoupperliftsinjection

 

Figure 4-7: Temperature change during air injection test 1

84

injection resulted in temperature increase in lifts 1, 2 and 3 at rates of 0.29i0.22 oC/day,

0.38:0.39 oC/day, and 0.15:0.14 oC/day, respectively (Figure 4-8). Lifts 4 and 5 had

negligible increase in temperature during the same period. On day 447, indicated by the
second dotted line in Figure 4-7, the air injection was shifted to lift 4. This shift in

location of air injection resulted in an increase in waste temperature in lifts 3, 4 and 5 at
an average rate of 0.07 °C/day, 0.6 °C/day, and 1.56 oC/day, respectively (Figure 4-8).
This shift in location also resulted in a decrease of temperature in lifts 1 and at an average
rate of -0.06 °C/day and -0.10 °C/day. This is due most likely to heat loss to the bottom

of the bioreactor landﬁll cell.

2.0

 

 

1.51

 

 

 

Rate of temperature change (C/day)
ES

[3612
0.0 . [Tr—1 [3041

Lift] Lift2 Lift3 Lirt4 Lifts

 

 

 

 

 

 

 

 

 

 

 

Figure 4-8: Rate of waste temperature increase during test 1

85

The temperature response during air injection test 2 is plotted in Figure 4-9 is a

representation of the same picture occurred in test 1, however for test 2. The temperature

increased in response to air injection through lift 1 at a rate of 0.21:0.10 °C/day,

0.23:0.11 °C/day, 0.08i0.03 °C/day, for lifts 1, 2, and 3 respectively (Figure 4-10).
Temperature increase in lifts 4 and 5 was negligible, with a calculated rate of 0.03:0.01

°C/day and 004230.01 °C/day, respectively. During the air injection through lift 4, the
waste temperature increased at a rate of 0.03i0.06 °C/day, 0.04:0.46 °C/day, 0.14:0.18
°C/day and 0.07i0.18 °C/day for lifts 2, 3, 4 and 5, respectively (Figure 4-11). The

temperature in lifts 1 decreased at rate of -0.05i0.01 oC/day.

When air injection was conducted through lift 3, the temperature increase in lifts 2, 3 and
4 increased at a rate of 0.06i0.05 °C/day, 0.43i0.32 °C/day, and 0.0532002 °C/day,

respectively (Figure 4-1 1).

86

 

 

40

35

30

25

N
0

Temperature (C)
G

 

 

 

 

10
5
0
-5
-10
650 700 750 800 850
Time (days)

[ Kites... .. Lift 5

L :1: Lift 4 1 Lift 3

I Lift 2 —Liﬂ 1

— Stop air injection — - - Start air injection

- :7-78witchto upper lifts injection

 

 

900

 

 

Figure 4-9: Temperature change during air injection test 2

87

45D
3'
NEW

:Ai

E: swag bu misc? 3E3 N 58 macaw 080.83 EEEomEB 833 no 88 omSo>< 57v o.— 5:—

 

 

 

 

(map/3) aﬁueqo ammrodwof, 30 91123

88

23B
NED
m8.
«3”

5.

V 9a n ma: Amsohu wage? 233 N “we macaw 83.8.: 0538983 883 we 88 owﬁo>< A _-v 0.59,.—

SE-

  
   

m .5..— v #5 m «E

 

 

I ..

 

 

do

(ﬂap/3) aﬁueqo armerodtuai 50 91123

89

4.1.3. Effect of moisture on rate of temperature increase

Variation in moisture content was expected to be a major factor impacting the rate of
temperature increase in response to air injection. Lifts 4 and 5 had lower moisture
content during air injection test 1 than air injection test 2. In order to increase the
moisture content in lifts 4 and 5 leachate was injected in the period between tests 1 and 2.
The cumulative amount of leachate recirculated before air injection test 1 and air
injection test 2 is shown in Figure 4-12. Waste with lower moisture content has lower
heat capacity leading to higher increase in temperature. The resulted in enhanced rate of
temperature increase in lifts 4 and 5 during air injection test 1 compared to that observed

during test 2. The rates of waste temperature increase during test 1 for lifts 4 and 5 were

1.56 °C/day and 0.07:0.18 °C/day, respectively. The rate of temperature increase during

test 2 was 0.60 °C/day and 0.14 °C/day for lifts 4 and 5, respectively (Figure 4-13).

Lifts 1, 2 and 3 had the same moisture content during tests 1 and 2. As a result, the rate
of waste temperature increase did not differ significantly between tests 1 and 2 (Figure 4-
14). Lifts 4 and 5 had lower moisture content in test 1 than test 2, thus leading to
enhanced rate of waste temperature increase during air injection test 1. Moreover, lifts l,
2 and 3 had the same moisture content during tests 1 and 2; therefore the rate of waste

temperature increase did not differ for these lifts.

9O

 

600,000 I
1 500,000

 

 

T

—4

400,000 j
300,000 -._
200,000 +1
100,000 “ﬂ

 

 

 

 

 

 

 

 

 

Leachate Injection (gal)

 

 

1 0 . .
l First Second Third Forth Fifth Sixth
. Lifts !

 

 

l D Befbre air im'ection 1 Before air injection 2 i i

Figure 4-12: Comparison of the leachate injection volume between tests 1 and 2

 

 

 

 

 

 

 

 

 

 

2.0
3‘
'3
$3. 1.5l
0
DD
5
4:
O
2 1.0.
*3
a
Q.
E
8
«5 .5'
8 _
m .A
M .Testl
0.0 .—, f—j—1 ‘ I “—1 [:]Test2

 

Lift2 Liﬁ3 -Liﬂ4 Lifts

Figure 4-13: Comparison of rates of waste temperature increase between air injection
testl and test 2 when injecting air through lift 4

91

 

 

 

 

9.
g .61
B
“33 .54
2
U
.E .4
0
E
g .3
d.)
o.
E. .2
“-4
o
2 .l .Testl
0.0 _ _ _ [:lTestZ
= 4 14 4 6 4 14
Liﬁl Lift2 Liﬁ3

Figure 4-14: Comparison of the rate of waste temperature increase between test 1 and
test 2 when injecting air through lift 1

4.1.4. Temperature response after air injection tests

Waste temperature decreased after stopping air injection for both air injection tests. A
decrease in the waste mass is attributed to injection of colder leachate than the waste
mass (temperature of injected leachate was 9.8 °C), heat loss to the bottom of the
bioreactor landﬁll, and heat conduction within the waste mass. Heat losses to the surface
of the landﬁll were eliminated because waste temperatures in lifts 4 and 5 did not
decrease when leachate recirculation was not taking place regardless of ambient

temperature (Figure 4-15).

 

 

 

 

 

 

 

 

Temperature (C)

 

 

 

 

 

 

‘ 0 50 100 150 200 250

 

"4 Air tenperature .2004 lift 5 . 2004 m" 4’ ‘

_l

 

 

Figure 4-15: Waste and air temperature after stopping second air injection test

93

Average temperature on the day of stopping air injection test 1 was 19.5:t3.0 °C, within
213 days the temperature decreased to 14.5:3.7 °C. After air injection test 2 the
temperature decreased from 28.2:3.6 °C to 22.1:22 °C after 232 days (Figure 4-16). The
rate of the temperature change was studied on the temperature of the landﬁll as a whole,
the temperature after both tests decreased, and the rate of decrease after tests 1 and 2 was

calculated to be -0.017 °C/day and -0.029 °C/day, respectively (Figure 4-16).

 

 

 

 

‘ 35 yet -0.0171x #2152745 y =’ -0-0239X + 53-167

i L R2 = 0.8865 R2 = 0.9308
1 30

 

 

 

 

N
M

 

 

Mean temperature (C)
N
o

 

 

 

 

 

; wt : : ~ ~ ~~~——-——~
' i I I I:

a‘ 10 .W . W Y" 1 W ___T_________,__ ..

. 400 600 800 1000 1200
1; Days

i i 0 After air injection 2 I After air injection event 1]!
i i- - - - Leachate injection — - - Smnp removed

 

 

 

 

Figure 4-16: Average temperature in the landﬁll after stopping air injection tests 1 and 2

Moreover, the variability of the rate of waste temperature change was calculated for each
individual location in the landﬁll to study if there is conduction in heat within the waste

mass. The variability in the direction of temperature change in the waste indicated that

94

some of the heat conduction was taking place within the waste mass. Some of the
measured locations had temperature increase while others showed a decrease in

temperature.

The largest decrease in temperature after air injection test 1 was observed for lifts 4 and 5
in which leachate recirculation was practiced (Figure 4-17). Waste temperature also

decreased in the lift underneath the lift receiving leachate (Figure 4-18).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 Inject 12,100 gal in lift 5 133011313 93°F“, ,
1 45 injection inliﬂs __ injection In hit 4 ﬂ _
1 u 5 and 6 i
; 40 ,_ - -- ---— - -_..
I | I I I
| . 35 | l I
l Q I I I
e 5 I
J 5 I I I
i i .
2%
I E— -
i 10 i l l l l
‘ hi I I I I
ij—M_ _ —— I I I
O LWW__. _W I I____ l 1 WI I
-2 48 98 148 198
Days
1 IWWW WWW, _ !
L 1 o Lrit 4 A Lift 5 - - - Leachate 'njection period}

 

 

 

Figure 4-17: Average temperature in lifts 4 and 5 after stopping air injection test 1 with
leachate recirculation periods marked

95 “‘

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1Q 401“ — —_ #— —— #* I I ”__“w—

E 30 _.__ — e I I ~ A“

e l I I

E“ 20 k'h‘w —J

B . - _

8 10" I I

a I l

3 O ‘T_‘—#" '_“ —T C“ ‘ l l +— 'T 1

0 50 100 150 200 250
Days from stopping air injection ‘

 

- - - - Leachate injection 136de in lift 4 l !

 

r—_._f22 __.. _._

Figure 4-18: Average temperature in lift 3 after stopping air injection test 1 with leachate
recirculation period through lift 4 marked

96

4.1.5. Heat Generation Factor

Input parameters and operational conditions

The heat balance was conducted on the bioreactor landﬁll using data from 13 air injection
events. Table 4-5 shows the operational conditions for each air injection event. These
include the duration of air injection, air injection flow rate, average air temperature,
average relative humidity, average waste temperature, and average oxygen concentration
leaving the landﬁll during each air injection. Table 4-6 shows calculation results of
input parameters that varied for each air injection. These include vapor pressure, oxygen
consumption rate, and rate of temperature increase in the landﬁll. Vapor pressure at
waste and air temperatures was obtained from the CRC handbook of chemistry and
physics [110]. The amount of oxygen consumed was calculated by applying equation (3-
31). The rate of temperature increase was equal to the slope of temperature vs. time
curve shown in Figure 4-19. Table 4-7 shows the constant input parameters for the heat
balance equation. All input parameters shown in Table 4-7 were either measured or
obtained from chemical and physical tables [110], except for the temperature underneath
the landﬁll and the heat transfer coefﬁcient. Discussion on estimation of these

parameters is described below.

97

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

N.N ﬁNw Nde wNN md v.2 §N\N\N~ vooQ _ C _ _ N3 2
02 ".3. mwmdd v.mN dd 0.5 .3de (a 383R :4 Nd
9M: m.Nm ocNdd EMN m d Nd vooQNE XENENE N5 ~ ~
03 93 chdd wNN 5m m voonQw 3dN\d:w N5 3
mad 04K w _ mod YNN Nd m— RENE :w RENEE m S a
Nd_ mdc ENdd N. _N Nd m: 3dN\m\w vodeNR 34 m
0.3 ddh moNdd wd~ _d E 33% :5 vodNNR 34 N.
E; wdc dedd 03 dd 5 vodNNR §N\mN\o m 3 c
92 QNN nomod 0.2 d d w 3oN\mN\c .3de to N: m
5.2 mNu 38d QE ﬁN MN vodNRCc 33ka :4 v

Md ddh domed QB wd NN mdthsNBﬁ modN\N\o~ :5 m

md v. S Nde 02 dd _.h mddNRNR mddNNNR N~O N
0.5 ".2. w :dd d: _d ﬁw mddQER mddNER N #0 ﬂ

Ems
Gov Aﬁb Shmv Gov “scum“ «a 333 8.5 3.5 35 5393.5
2522—88 3:55... 8a.. 23.22.88 net-559:8 562:9 dam 25m cowoo E a?
:4 9523— ”a.“ 33>? 5»sz “2

 

 

Eogogm 385830 :05 can 36358 mm? ooze—an «no: 05 :02? 8m 3:26 8302.5 5w ”m... 035.

98

Table 4-6: Calculation results of input parameters that varied for each air injection

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Vapor pressure Vapor pressure 0 Rate of
Air at arr at waste xygen temperature

injection temperature temperature consumed increase

(10'2atm) (10"atm) (“‘"WS’ (°C/day)
1 1.99 1.30 0.29 0.15
2 1.17 1.49 0.43 0.33
3 1.08 1.90 0.28 0.18
4 1.76 1.78 0.17 0.14
5 1.68 1.99 0.30 0.07
6 1.92 2.08 0.31 0.13
7 2.12 2.27 0.26 0.11
8 2.20 2.48 0.26 0.07
9 1.77 2.68 0.31 0.12
10 1.72 3.69 0.23 0.04
11 2.11 2.90 0.25 0.05
12 2.10 3.21 0.55 0.18
13 0.71 3.70 0.42 0.02

 

 

 

99

Table 4-7: Input parameters for the calculation of heat balance equation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Parameter Value Unit
The landﬁll footprint area (A) 4856 m2
Waste Density (P) 599.21 kg/m3
Waste volume (V) 41,330 m3
Speciﬁc heat of water 4.1818 J/ kg °C
Speciﬁc heat of solid fraction of waste 1.2 J/ kg °C
Speciﬁc heat of wet waste (Cp) 2,491 J/ kg °C
Speciﬁc heat of dry air (Cpair) 1,030 J/ kg °C
density of dry air 1.2929 kym3
R (gas constant) 0.082057 atm L/mol K
Heat of vaporization for water 2.444667 kJ/g
Pressure in the landﬁll 1 atm
Heat transfer coefﬁcient (U) 3.18E-04 kJ/(s m2 °C)
Temperature beneath the landﬁll 11 °C
Moisture content 43.3 %

 

100

 

 

25 _._... --——~ ---——— «-
‘ y=0.1221x- 70.604

R2 = 0.9965

 

i
i

 

 

N
W

 
 

 

 

Waste temperature (C)
N
N
T— _—

 

 

 

 

 

 

 

754 756 758 760 762 764 766 768 770
Days

 

 

 

Figure 4-19: Determination of the rate of temperature increase in the bioreactor landﬁll
for air injection event 9

Temperature of groundwater in Michigan was used to represent the temperature of soil
underneath the landﬁll, because the temperature of soil underneath the landﬁll could not

be measured directly. Groundwater nearby the surface is at constant temperature during
the year [124]. The average shallow groundwater temperature in Michigan is 11 °C
[125]. A study conducted in Oakland County in Michigan concluded that groundwater

temperature samples ranged from 10.4 °C to 15.5 °C, with a mean of 12 °C [126]. The
temperature of ground water is usually 1 to 2 oC higher than the mean annual air

temperature [127]. The mean air temperature in Harrison in 1993 and 1994 was 6.7 °C

101

as measured at the weather station in Northern Oaks bioreactor landﬁll. This results in a

groundwater temperature of about 8 °C as described by Todd [127]. As a choice of input

parameter, temperature of 11 °C was used to represent the soil temperature underneath

the landﬁll.

Heat transfer coefﬁcient was calculated for the bioreactor landﬁll by conducting heat
balance after the second air injection test was stopped. During this period neither air nor
leachate were recirculated in the landﬁll. Therefore the value of latent, sensible, and
metabolic heat can be taken as zero. Substituting in equation (3-22), the heat balance

equation after the second air injection test is given by equation (4-2).

pCpV % (T) = UA(T0u, — Twang) ............................................................................. (4-2)

Rearranging, equation (4-2), we get:

dT
= pVCp /dt
A(T out ’ T waste)

 

................................................................................................ (4-3)

Where,

dT/dt is the rate of temperature decrease in the landﬁll after the second air injection test

and equals 0.0289 °C/day given by Figure 4-16, and

Twaste average waste temperature during this period equals to 24.36 °C.

102

The computed value of heat transfer coefﬁcient was 3.18 10'4 kJ/(s m2 °C).

Heat balance analysis

Dynamic calorimetric method was used to calculate the heat generation factor of waste in
this study. This study represents the ﬁrst reported effort to use landﬁll as a calorimetry to

determine heat generation factor. The only other study reporting the use of larger size

reactors than laboratory scale reactors, used a size of 21.4 m3 in 1856, 150 years ago

[128] cited in [55]. Another study indicated the use of commercial size to determine heat
generation, but they did not specify reactors making the comparison with difﬁcult [55].
The size 21.4 m3 is 1,931 times less than the bioreactor cell. Field scale evaluation of

heat generation factor are more valuable than those obtained using lab scale reactors,

because it reﬂects the amount of heat generated under ﬁeld conditions.

The results of the heat balance analysis for each air injection test are presented in Table

4-8. This table includes all ﬁve terms in the heat balance equation and the value obtained

for the heat generation factor. Heat generation factor ranged from 200 kJ/mole 02 to

626 kJ/mole 02 with an average of 354188 kJ/mole Oz. The range observed in this study

is larger than the range observed for single substrates under laboratory conditions (397 to
543 kJ/mole 02) [55]. The range of heat generation factor calculated for Northern Oaks
bioreactor landﬁll was within the same range of that observed for single substrates under
laboratory conditions (Table 4-9). To perform this analysis, 29 reported heat generation
factors were obtained from a study that compiled the heat generation factor in literature

[55] and a t-test was conducted to compare these values with those calculated in this

103

study (4-10). A t-test indicated that the average obtained under ﬁeld conditions for solid
waste is signiﬁcantly less than that obtained under laboratory conditions for single
substrates. Although the heat generation factor for single substrates (29 measurements)
was found to be high compared to the heat generation factor values determined in the
bioreactor landﬁll, many more ﬁeld studies measurements spread over different
bioreactor landﬁll must be performed to conclude that single substrate values are always
higher than complex substrate measurements. Indeed, in one study the heat generation
factor for a synthetic solid waste was determined under laboratory conditions to be 304
kJ/mole [24]. If the value determined for solid waste under ﬁeld conditions is truly lower
than that observed for single substrates then the heat generation factor of single substrates
under laboratory conditions, then using the single substrate valued would be an
overestimation of the amount of heat produced for each mole of oxygen consumed.
Studies that have focused on heat transfer and modeling in landﬁll environments have

used values for single substrates [23, 58].

104

Table 4-8: Heat loss, production and accumulation in the bioreactor landﬁll

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Ai -HGF Heat Metabolic Sensible Heattoloss L133“
injectrion (ET/mole) accumulation beat heat due boundary loss
(kJ/s) (kJ/s) (kJ/s) (kJ/s) (kJ/s)
1 370 109 109 0.36 -0.12 0.16
2 559 237 241 -0.23 -3.60 —0.46
3 490 127 137 -O.44 -9.07 -0.80
4 627 99 107 -0.01 -7.58 -O.20
5 200 48 59 -O.1 1 -10.27 -O.46
6 328 89 101 -0.05 -11.28 -O.52
7 349 76 89 -0.04 -13.54 -0.31
8 237 47 63 -0.07 -15.71 -O.52
9 345 87 106 -0.28 -17 .65 -0.86
10 245 29 57 -0.48 -25 .96 -1.40
11 220 34 55 -O.18 -19.71 -0.59
12 275 126 150 -O.54 -22.30 -1.88
13 191 13 81 -1.51 -26.02 -2.54

(-3736‘)

 

* Sensible heat loss due to leachate recirculation

105

 

Table 4—9: Average heat generation factor

 

 

 

 

 

Std.
Heat generation factor Std.
(kJ/mole) N Mean Deviation Error
Mean
Literature data for single ‘
substrates [55] 29 445.6 37.2 6.9
Synthetic solid waste
[24] 1 304
This study 13 341.2 139.6 38.7

 

 

Table 4-10: Comparison between laboratory values for heat generation factor in
laboratory (single substrates) and ﬁeld scale (solid waste).

 

 

 

Si (2_ Mean Std. 95% Confidence
t (if “Sic d) Difference Error Interval of the
Difference Difference
Lower Upper
2.65 12.8 0.02 104.3 39.34 19.2 189.5

 

 

 

Table 4-11 shows total heat loss and heat gain and the percent contribution of each term
in the heat balance equation. Heat accumulation ranged from 51 % to 100 % of the
produced heat when no leachate recirculation was taking place. With leachate
recirculation, the heat accumulation was 17 % of the total heat produced. This indicates
the strong effect of leachate recirculation on the temperature of the waste, particularly

when the leachate temperature is low. The temperature of the recirculated leachate was

10 oC. Fifty percent of the heat loss during leachate recirculation was sensible heat loss.

106

The above data indicate that leachate recirculation can be used as a temperature control

mean to minimize ﬁre hazards during air injection.

Sensible heat due to air injection was minimal and ranged between zero percent and 5.3
%. This term is much lower than found in commercial size reactors, where 9-12 % of the

total heat loss was sensible heat [58].

The latent heat is usually assumed to be the biggest mechanism of heat loss in larger
reactors, reaching 80 % of the total heat loss [58]. In this study it ranged between 0 %
and 10.7 %. This low percentage of heat loss can be attributed to the high relative

humidity in the injected air.

Heat balance studies on relatively big reactors with high volume to surface ratio, usually
ignore the heat loss to the boundaries [24, 55, 58]. In this study, it was found that the
heat loss to the bottom of the landﬁll was signiﬁcant. Under no leachate recirculation
conditions the heat loss to the boundaries ranged from 84 % to 100 % of the total heat

loss (Table 4-11).

107

8.3388. 228.... e 8.. a... .8. ”3.80m ..

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Came
N.N w.m 9mm 2 E #60. 2
N.N 0.5 Nda V» 02 WVN- N—
md QN Ném no mm WON- u w
NA 06 m.m¢ _m hm wﬁm- A:
WA 9? ada mm 9: w.w~- m
v.0 N.m #60 E. mw md—i w
Md m.N 155 Va aw Q2- 5
m6 vi Nda mm 2: m.:- @
o4 m6 Eva mm mm wd—t m
ﬁo WN vﬁa ma 2: QB. v
m4... mg. mﬁm Na wm— WOT m
Md 5A: 64% ma ~VN m4». N
Cd od odo— co“ 9: ﬁc- m
Ae\ev
AX; ax; 68.69... 3...: 3...:
.é a... .88 a...» .8 5.8.3.
m8. .3.. 2.3.5 b.652— ue 2.3.2— 2. . _ .._<
.3 o .28 e mm:
a .— 3. m a ‘— 23.. «no: «no:
132......83.
an... .38 he 2.8.8: a as mne— «no:

 

 

nouns—co 02.2.3 .8.— oﬁ mo 8.8. some .«0 8:35.80 2.8.3 on. v...» 5mm EB 32 .3: :33. u. 7.. 033—.

108

4.1.6. Change in oxygen, nitrogen, carbon dioxide and methane concentration due
to air injection

The gas phase in anaerobic bioreactor landﬁll is dominated by carbon dioxide and
methane. Air injection into the bioreactor landﬁll leads to the change in this gas phase
composition at the affected locations. Figure 4-20 is an example of gas phase
concentration change at a single sampling port in the landﬁll during the air injection
experiments. Air injection leads to an increased oxygen and nitrogen concentrations and
decreased methane concentrations in the landﬁll. The injected air penetrates into the
landﬁll pores, replacing both carbon dioxide and methane. Over time, it reaches to a
relatively stable condition, at which change in concentrations is minimal as long as air
injection is still in operation. The process of replacement is very quick; within a few
hours the concentration of methane and carbon dioxide decreases and the oxygen and
nitrogen concentration increases. Initially the rate of change is faster which slows down
gradually. The time required to reach the steady state condition was not monitored
closely, however it was seen that after 4 days of aeration, the concentration of nitrogen
reaches a stable value. The concentration of the gas components after 4 days of aeration
was similar to that after 11 days. Figures 4-20 and 4-21 show speciﬁc examples of this
process at a location that is 20 ft away from the air injection line. Figure 4-21 shows that
40 minutes prior to the start of air injection, the concentration of carbon dioxide,
methane, nitrogen, and oxygen were about 65 %, 30 %, 10 % and zero, respectively.
Within 5 hours of starting air injection the oxygen and nitrogen increased to about 10 %

and 40 %, while methane and carbon dioxide concentrations were reduced to 10 % and

109

40 % (Figure 4-21). As time progressed (Figure 4-24) these concentration changed
further, towards higher oxygen and nitrogen concentrations, reaching a state of ‘steady
state’, where change became minimal. The close to zero methane concentration and 5%
carbon dioxide concentration indicates that ﬂushing of the gas phase was taking place.
Most of the methane produced and accumulated in the locations exposed to air was

ﬂushed out by the injected air. Other ports (15) exposed to air showed similar.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Time (days)

{—9-CH4 +c02 +02 +N2

 

 

i 80
1 70
1 A60 ~
@150 .
8
2:40 -_
E
830
c:
152° 4><'
1 1. N
1 \-
' O l I a __.-————
i 0 l 2 3 4 11
1
1

 

 

 

Figure 4-20: Change in gas phase concentration during air injection at location G211

110

 

OIL
0U

 

 

 

 

 

 

 

 

 

 

 

 

 

1,. “I: :79 ‘rStartairirjection _ N2 _-_.
13 1. 4414—4 —

:§ “_50 (:02 ﬁll—_l 4

.. 4 >—
15 - 4 +
. o

: G

O

I)

 

 

 

 

 

 

 

 

0.00 0. 04 0.09 0.14 0.19 0.24
Time (days)
1 15.79114 +C02 +02 +102]

 

 

 

 

Figure 4-21 Change in gas phase concentration during the initial hours of air injection at
location G211

4.1.7. Impact of air injection on leachate characteristics

Laboratory and ﬁeld scale bioreactor landﬁlls have demonstrated that leachate can be
treated by recirculation. The process of cleaning leachate aerobically is faster than that of
anaerobic process at optimum conditions, because of the faster degradation rates. At
Northern Oaks bioreactor landﬁll, the pH, COD and VOC of leachate was studied within
the waste using sampling ports and at the sump. Leachate samples from the sump were
collected monthly, but from the sampling ports it was collected less frequently for
analysis. Leachate samples from sampling ports were collected every three months and
before and after air injection. However, during the air injection test 2, leachate samples

were collected on a weekly basis. Leachate sampling from some of the ports was not

111

possible due to lack of sufﬁcient leachate in the sampling basin. On an average only 20 to

35 ports were usable out of the 48 ports.

Before the air injection experiments, very high concentrations of COD and VOC were
observed that was accompanied by low pH in all ports. After air injection, COD and
VOC dropped rapidly and pH increased. Figure 4-22 shows the average concentration of
COD in lift 1 for 13 sampling ports before and after the two air injection tests (i.e., before
air injection test 1 and after air injection test 2). The average COD concentration for lift
1 before air injection test 1 was seven times higher than that after air injection test 2, a
decrease from 21 g/L to 2.9 g/L. Similar decrease in COD occurred in lift 3 (Figure 4-23;
three separate ports). The third port did not have enough leachate in the beginning. The
ﬁrst sample obtained was 20 days after air injection, which was ~ 23 g/L. It decreased to
3 g/L after air injection. More than 90% reduction in leachate COD was observed at most

ports.

The impact of air injection and decrease in COD on pH was expected. Figure 4-24
depicts the correlation between pH and COD before and after air injection. Before air
injection, the pH was in the range of 5 .3 to 6.3, and the average COD was ~34 g/L. After

air injection, the pH was in the range of 7.2 to 8.3, while average COD was about 3.2 g/L.

VOC concentrations were also measured in leachate, although less frequently than other
parameters. The following VOCs were detected in the leachate: acetone, methyl chloride,
methyl ethyl ketone (MEK), methyl isobutyl ketone (MIK), toluene, and xylene. MEK
and acetone had the highest concentrations and methyl chloride and xylene had the

lowest concentrations among all the measured VOCs. Table 4-12 lists the VOCs and

112

their concentrations for two sampling events, one before the air injection and one after.
As with COD, the concentration of VOCs reduced signiﬁcantly after air injection.
However, the sampling events for VOCs also included the impact of losses due to
volatilization and aeration over a period of one year. During this period, air injection was

in place and lift 1 was exposed to air for more than ﬁve months.

Table 4-12: Average concentrations of VOC in lift 1

 

Methyl
chloride MEK MIK Toluene Xylene

(Hg/L) (pg/L) (113/L) (us/L) (pg/L) (Hg/L)

Acetone

 

9/2/2003 24,681i13,685 79i87 34,948i26,555 522:1:498 108:1:60 41:4

 

 

10/15/2004 1,252i1,670 0 301-59 5:17 31i26 4i5

 

 

 

 

 

1 40 : 4444 ----- «
1 ’ 4 444 4 4 -— COD before air injection —444. ..
1

I 241- 31 - 444— coo afterairinejction

/ -

Miriam a 2.897

 

b)
(It

 

 

Thousands
N b.)
u. o

 

 

 

p—n
M

 

Average COD (mg/L)
N
o

 

 

 

 

 

—v

 

...-A
OUIO
I
K
V

650 700 750 800 850 900 950 1000

 

Figure 4-22: COD concentration before and alter air injection in location lift 1

113

 

 

 

 

 

 

 

 

 

 

 

 

A -8 60 I A "*ﬂ i
E g 50 " ‘“’ Y 1
V .1: N A
g 30 1— -— 4 - __..
0
g 20 _l__ “_.__ﬁ __ __
a |
O 10 7*“; -—-—--- - .3
U | K;

0 “i“ ‘ ' 1 7 1 1 1 1

 

 

 

Figure 4-23: COD concentrations decrease during air injection in lift 3.

 

 

 

 

 

 

 

100,000 444 20,000
.1: ‘
i: L o e 16000.5
,2 A ‘ —¢-—¥—- 34,387 ; 4 g .
..D
0 ﬁ ' O a 3
g g 1 9 412,000,, g
g g 10,000 1 E g
. a .
§ 1,; 1 4 8,000 ,3; g
.. 1
Q 1 I "f 4,000 O
8 ‘ F. 3,180 8
1,000 4.4—4444 4 - 4 4 - 0
5.00 6.00 7.00 8.00 9.00
pH
*7 Beforegﬁijﬁectioﬁ ‘ __ Average 9 I _ I Aﬂerairinjecﬁon _ l,

 

 

 

 

 

 

Figure 4-24: COD concentration vs. pH before and after air injection in sampling
location W111 — Lift 1

114

 

 

Figures 4-25 and 4-26 show the distribution of acetone concentrations within lift 1.
Figures 4-25 show that the concentration before air injection (2003) was in the range of
7-42 mg/L, with an average of 25 mg/L. In 2004, acetone concentration decreased to
below 2 mg/L (Figure 4-26). It was interesting to note that the port at which acetone was
still present after one year, were the locations where it was low in 2003 indicating the
role of air injection as a major mechanism of removal. The role of air injection as a
major mechanism of removal was studied by calculating the resistance to mass transfer
from the liquid and gas phase for the VOCS found in leachate in Northern Oaks
bioreactor landﬁll To determine which phase controls the mass transfer, the following

equation was used [129]:

 

gim) = 1001 ...................................................... , .................................... (44)
T 1 +
(kg / k1 j;
Where,

RL is the liquid phase resistance to mass transfer,

RT is the overall resistance to mass transfer,

k3 is the gas-phase mass transfer coefﬁcient that describes the rate at which contaminant

A is transferred from the air—water interface to the bulk gas phase,

115

k1 is the liquid-phase mass transfer coefﬁcient that describes the rate at which

contaminant A is transferred from the bulk aqueous phase to the air-water interface, and

H is Henry’s constant in (Lliquid/LAjr ).

The ratio (kg/k1) ranges between 40 to 200 depending upon the type of aeration or

stripping system used [129]. For a ratio (kg/kl) of 100, RL/RT was calculated by equation

(4-4), using Henry’s constants given in Table 4-13. The results of the calculation are
shown in Table 4-14. For MEK and MIK the mass transfer is controlled by the gas phase
only. For methylene chloride, toluene, and xylene the mass transfer is controlled by both
phases, but predominantly by the gas phase. With temperature increase the contribution
of resistance from liquid phase becomes closer to that of the gas phase. For acetone mass
transfer is controlled mainly by the gas phase, as temperature increase the contribution of
the liquid phase increases slightly. The volume of injected air was 17 times the volume
of the waste. All of theses chemicals are easily volatile and some are also easily
biodegradable such as acetone. Therefore, it is possible that all chemicals were
volatilized before experiencing any signiﬁcant degradation however for acetone a

combined mechanism for loss is more likely.

116

 

2604.00

 

 

 

 

 

 

 

 

 

42000
36760
31500
2556.00 26250
21000
15750
10500
5250
2828.001111] 0
200.0000] 2506.00 2534.00 2502.00 2590.00
Figure 4-25: Acetone concentrations before air injection in lift 1 (9/2/2003)
Between 2,600 and 5600 jig/L
2604.00
+
+ +
2656.00
4‘ 4»
Less than 2,600 ng/L
+
252000011]
2478.00[m] 2508.00 2534.00 2562.00

 

Figure 4- 26: Acetone concentrations after air injection in lift 1 (10/15/2004)

117

 

Table 4-13: Henry’s constant for VOCs found in leachate samples from Northern Oaks
bioreactor landﬁll [129]

 

Henry’s constant

 

 

 

 

Name
(Lamar/Lair)

Temperature 10°C 20 °C 25 °C 30°C
Acetone‘ 8.4310“4 1.3010“3 1.5910‘3 1.9110‘3
MEK 2.80104 1.9010‘l 1.30104 1.1010’4
MIK 6.60104 2.90104 3.9010‘4 6.8010'4

 

Methylene chloride 1.4010"3 2.4410”3 2.9610'3 3.6110“3

 

 

 

 

Toluene 3.8110‘3 5.5510‘3 6.4210‘3 8.0810’3
-3 -3 -3 -3
m-Xylene 4.11 10 5.98 10 7.4410 8.87 10
-3 -3 -3 -3
p-Xylene 4.2010 6.45 10 7.4410 9.45 10
-3 -3 -3 -3
o-Xylene 2.8510 4.74 10 4.8710 6.2610

 

 

 

*Source for acetone data: http://www.epa.gov/athens/learn2model/part-
two/onsite/esthenry.htm

118

 

 

Table 4-14 Ratio of RL/RT VOCs found in leachate samples from Northern Oaks
bioreactor landﬁll for kg/kl=100:

 

 

 

 

 

 

 

 

 

 

 

 

Name RL/RT (%)
Temperature 10 °C 20 °C 25 °C 30 °C
Acetone 8 12 14 16
MEK 3 2 1 1
MIK 6 3 4 6
Methylene chloride 12 20 23 27
Toluene 28 36 39 45
m-xylene 29 37 43 47
p-xylene 30 39 43 49
o-xylene 22 32 33 38

 

4.2. Oxygen Consumption Rate

Attaining optimal temperature of waste is an essential requirement for enhanced
biodegradation. This is especially true for colder climates when the waste is placed at
lower ambient temperatures. The fastest and most economical method to increase
temperature in a landﬁll is by air injection. Air injection creates aerobic conditions in the
waste resulting in exothermic reactions. The rate of aerobic microbial degradation affects
the size of the aerobic zone for a given air injection ﬂow rate. Higher oxygen
consumption rates requires higher air flow rate. To properly design an air injection

system, the rate of oxygen consumption must be known. This is accomplished in

119

laboratory scale reactors under highly controlled conditions. Such conditions cannot be
achieved in a bioreactor cell. The difference in conditions and scale may lead to
difference in the oxygen consumption rate, because of heterogeneity of waste. In this
section oxygen consumption rate inside the landﬁll and at vents are described. The
effect of background oxygen concentration, moisture content, and temperature on this

rate is discussed in this section.

4.2.1. Operational conditions

Oxygen vs. time curves were obtained at the sampling locations twice, separated by 7
days of cumulative air injection period during the year of 2003. The two tests were apart
by 7 days of active air injection. The sampling locations at which POCR was determined
in the ﬁrst and second tests are shown in Tables 4-15 and 4-16, respectively. These
tables describe the oxygen concentration before stopping air injection, moisture content,

and waste temperature during the period of POCR determination for each location. The
temperature at locations in Test 1 ranged between 3.7 and 24.7 °C, volumetric moisture

content ranged between 0.31 and 0.65, and starting oxygen concentration ranged between

4.4 and 19.8%. The ports sampled in the second test had a range of temperature, moisture
content and starting oxygen concentrations of 10.5-51. 0°C, 0.35-0.66, and 2.2-20.3%,

respectively (Table 4-16). Together, these experiments provided the data for studying the

effect of temperature, moisture and starting oxygen concentration on POCR.

Electronic Conductivity in Leachate samples was measured before and after the air

injection, an average of the two samples was used. At locations were leachate samples

120

were not obtained, the average of the leachate value for the lift from which the location

was measured was used.

Table 4-15: Sampling locations, temperature, moisture content and oxygen
concentration before stopping air injection at which POCR was conducted (Test 1)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Average
Location temperature Moisture Starting oxygen
content concentration (%)
(° C)

9/16/03
G132 14.8 0.46 4.4
G143 16.7 0.53 12.5
G211 16.7 0.64 17.8
G213 14.0 0.60 13.5
G214 13.3 0.65 9.1
G221 23.7 0.45 19.7
G222 13.4 0.60 17.9
6223 18.1 0.60 16.7
G224 20.1 0.60 14.9
G231 11.9 0.64 19.8
G321 11.6 0.46 12.2
G322 10.5 0.57 12.4

10/17/03

G512 6.2 0.31 19.6

G521 3.7 0.48 19.2

G522 7.4 0.51 18.7

 

 

121

Table 4-16: Sampling locations, temperature, moisture content and oxygen concentration
before stopping air injection at which POCR was conducted (Test 2)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Location Average temperature Moisture content Starting oxygen
(°C) concentration (%)
9/29/03
G123 10.5 0.52 2.2
G213 15.8 0.54 11.5
0214 16.7 0.66 8.0
G221 51.0 0.54 20.3
G222 20.3 0.54 10.0
G223 21.2 0.54 15.4
G224 24.8 0.54 13.1
G225 19.7 0.54 2.7
G231 15.6 0.65 20
G233 15.2 0.53 8.9
G321 15.8 0.45 1.6
G322 12.4 0.57 7.2
10/24/03
G512 39.2 0.35 13.9
G522 18.4 0.49 10.8

 

 

4.2.2. Fitting POCR data to Monod equation

Liquid phase oxygen concentrations over a range of approximately 8 hours were ﬁtted to
Monod equation. An example of the ﬁtted values for sampling location G221 is given
below. For this sampling location, the parameters estimated for k and Ks with 95%
conﬁdence interval were 4.76:0.02 per h, and 4.43i0.0003 mg/L, respectively. Initial
substrate concentration was estimated to be 3.22i0.01 mg/L. Concentration in the gas

phase for this sampling location ranged between 0.1 and 21%. This range is the

122

maximum range that can be obtained when injecting air. A ﬁt was obtained (Figure 4-27)
with the minimization function of a value f=0.0047 (equation 3—42). A wide range of

initial values were used to avoid local minima, forty iterations performed using different

initial values of k, Ks and So.

 

 

 

 

 

 

 

3.5:

   

b.)
T-

1"
Ln

 

Oxygen concentration (mg/L)

 

 

 

 

 

 

 

 

Figure 4-27 Fitting Monod equation to the experimental data for location G221

123

The validity of these parameters in comparison to the requirements for their

determination is presented below. In order to obtain a unique solution the ratio So/Ks

should be bigger than 0.1, this requirement was met as the ratio obtained was 0.68. This
is supported by the results of the 40 iterations performed which gave standard error for k

and Ks of 0.5% and 0.007%, respectively. In order to obtain independent coefﬁcients,
the ratio So/Ks should be larger than 1. This requirement was not met as the ratio
calculated was 0.678. In order to obtain valid estimates of parameters, the range of the

substrate concentrations should in the three regions deﬁned by Monod equation as

described (Table 4-17).

Table 4-17: Comparison between the range of data used and the required to estimate
Monod kinetics parameters

 

Range of data measured Range of data required

 

Minimum Maximum RegionI RegionII Region III

 

s/ks 0.0370 1 .00 0-2 2-9 >9

 

V/k 0.0357 0.50 0—0.67 0.67-0.9 >0.9

 

 

 

For sampling location 6221, the rate of oxygen consumption was calculated based on the
estimated kinetic coefﬁcients using Monod equation (3-41). This rate is plotted as a
function of substrate concentration in Figure 4-28. In this ﬁgure the theoretical rate is
extended to cover the three regions as described above. The ﬁrst region is located at

concentration range from 0 -9mg/L (shown by the ﬁrst dotted line and labeled as region I

in which 0 < S/KS $2). The second region is in the range between 9mg/L and 40mg/L (2

< S/Ks S9). The third region is at concentrations higher than 40mg/L corresponding to

124

(S/Ks 29). The oxygen concentrations measured in the ﬁeld were all located within the
ﬁrst region located between 0-4.4 mg/L. From this analysis we conclude that the

requirement to cover the three regions in determining Monod equation coefficients was

not met (Table 4-17).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 Region] _*ﬁ _ ﬁ_ __— __—
4 4 _ ._.._--_._.__
l ————————— I I I I I I I
I T— . FH‘” ' ' n FT
9 " --..-- _ I -f _~
1 n. . _ ' I
l 1!____. _ __ r ____
, ‘—— -ul—-””
1 . i .
: Reglon II I Reglon III
. l
i _ -____.|__ __.
1 4 Oxygen =21% 4W4" 44—414444 >44 4
i , -_ l __
i 20 30 40 50 60 '
3 Oxygen Concentration (mg/L)

 

Figure 4-28 Monod equation based on the estimated parameters for location 6221, T =
50 °C, y=l.15

In conclusion Monod equation cannot be used to model POCR data. Although the
measured data appears to ﬁt Monod equation (Figure 4-27), the requirements to use this
model were not met for to be applicable. Moreover, the range of data required to obtain a
valid ﬁt for the data is impossible because this range extends beyond the saturated
oxygen concentration. The parameters estimated are dependent, and located within the

region Monod equation reduces to ﬁrst order kinetics.

125

4.2.3. Fitting first order kinetics to POCR data

POCR at all the locations were modeled successfully using ﬁrst order kinetics. An
example of the ﬁt is shown in Figure 4-29 for location 6221. In each test POCR was

determined for 15 locations within the landﬁll, of which 12 locations were studied in both

tests 1 and 2. The ﬁrst order kinetic for this location is 0.51 per h with R2 of 0.99. First

order POCR was calculated for tests 1 and 2 (Table 4-18). The mean POCR of the 15
sampling locations for tests 1 and 2 were 0.0867i0.0422 per h, and 0.2450i0.0843 per h,
respectively. To evaluate the POCR difference between the two tests Wilcoxon signed
ranks test was used. This test was used because POCR in test 1 was not normally
distributed (Table 4-19). The result of this test indicates that there is a signiﬁcant
difference between the median of the two tests (Table 4-20). In Wilcoxon signed ranks
test, median is a better representation of the data than mean. The medians of POCR in

tests 1 and 2 were 0.041 (per hr) and 0.23 (per h), respectively (Figure 4-30).

126

Table 4418: First Order POCR

 

First order POCR First order POCR

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Location Lift number for Test 1 for Test 2
(per h) (per h)

1 6123 Lift 1 0.106

2 6132 Lift 1 0.201

3 6143 Lift 1 0.250

4 6211 Lift 2 0.197

5 6213 Lift 2 0.017 0.038

6 6214 Lift 2 0.058 0.201

7 6221 Lift 2 0.046 0.510

8 6222 Lift 2 0.039 0.399

9 6223 Lift 2 0.022 0.219
10 6224 Lift 2 0.044 0.123
11 6225 Lift 2 0.146
12 6231 Lift 2 0.044 0.125
13 6233 Lift 2 0.206
14 6321 Lift 3 0.049 0.376
15 6322 Lift 3 0.041 0.122
16 6512 Lift 5 0.167 0.331
17 6521 Lift 5 0.042 0.238
18 6522 Lift 5 0.086 0.537

Number 15 15

 

127

 

Table 4-19: Tests of Normality for POCR, temperature, moisture content and
background oxygen concentration

 

Variable Kolmogorov-

 

 

 

 

 

 

Smirnov Shapiro-Wilk

D df Sig. w df Sig.

First order POCR(testl) 0.303 12 0.003‘ 0.706 12 0001"
First order POCR(test 2) 0.158 12 0.200 0.939 12 0.489
Temperature (test 1) 0.166 12 0.200 0.972 12 0.935
Temperature(test2) 0.262 12 0.022‘ 0.813 12 0.013‘
BaCkgm‘md °"t§f‘l‘)°°“°ent’ati°n( 0.174 12 0.200 0.882 12 0.075

 

Background oxygen concentration( 0 101 12 0 200 0 972 12 0 913

 

 

 

test Q
Moisture content (test 1) 0.227 12 0.088 0.876 12 0.079
Moisture content (test 2) -.233 12 0.072 0.925 12 0.334

 

* Signiﬁcant (i.e. does not satisfy normality assumption)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 1
1 A ‘41

1 '3}, 1

1 E 1

1 g _. 1

1 -a - x

1 E y=4.8745e0'51 _ 1

1 4g 1" R2=0.9916 _1

1 8 1“ “"1 1
1 =1 1 1 1
1 86 1 -1 . .-.--..-.. __-__.__.1

1 O 0 . , 1 1
1 0 2 4 6 8 1
1

1 Time (hr) 1
1 1

 

Figure 4-29 First Order POCR for location 6221

128

 

Table 4-20: Wilcoxon Signed Ranks Test for comparison between POCR tests 1 and 2

 

 

 

 

 

Tests value First order POCR Temperature
Z -3.059 -3.059
Asymptotic Signiﬁcance (2-tailed) 0.002 ' 0.002 .

 

* Signiﬁcant at 95% conﬁdence interval

 

 

 

 

 

 

 

 

 

 

 

 

.6

A .51

{—4

3

O

r: .41

13

81

at .31

O

0

ﬁ .21

Q) *

“U

lI-l

3 1‘ 0

LL 0.01
-.1 _ _
N= 12 12

Testl Test2

Figure 4-30: First order POCR for two tests separated by 7 days of air injection

4.2.4. Factors affecting POCR

In order to analyze the reason for difference in POCR between the two tests, differences
and similarities for initial oxygen concentration, moisture content and temperature are

described below. Wilcoxon signed ranks test was used to test the difference in

129

temperature. This test was used because temperature data during the second POCR test

was not normally distributed during the second test (Table 4-19). The median
temperature during test 1 was 12.6 °C and during test 2 was 17.5 °C (Figure 4-31). The

difference in median was signiﬁcant between the two tests (Table 4-20). Paired t—test
was used to study the difference of background oxygen concentration and volumetric
moisture content because they were normally distributed (Table 4-19). The volumetric
moisture content did not differ between the two tests (Table 4-21). The average
volumetric moisture content was 0.54:0.06 during test 1 and 0.53:0.05 during test 2
(Figure 4-32). Background oxygen concentration differed signiﬁcantly between the two
tests (Table 4421). Average background oxygen concentration of test 1 was 15.282t2.82%

and 11.96i3.12% for test 2 (Figure 4-33).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

60
504 *
6 40‘ o
d)
5
e 301
O)
E‘ —— _.—
g 201
101
0 w I
N: 12 12
Testl Test2

Figure 4-31: Temperature during the two tests separated by 7 days of air injection

130

Table 4-21: Paired samples t-test for comparison between tests 1 and 2

 

 

 

 

 

Tests value Moisture content Background 0.“ gen
concentration
1 0.747 2930
df 11 12
Signiﬁcance .. .
(2-tailed) 94“ 0.013

 

 

* Signiﬁcant at 95% conﬁdence interval, ** Not signiﬁcant

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.7

§ .61

c:

o

O

O

5 .51

.2

O

E

O

E .4

’8’

.2

9 31 _—
.2 - '
N= 12 12

Testl Test2

Figure 4-32: Volumetric moisture content during the two tests separated by 7 days of air
injection

131

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

g 30
0
8
J:
0..
26
00
o 201 __.— ——
E.
.E
C
.2
*6
E
8 10‘ _L__
1::
O
Q
t:
&
>~ __L_
5 0
N= 12 12

Test 1 Test 2

Figure 4-33: Oxygen concentration before turning off the air injection

Although POCR did not correlate with temperature, oxygen concentration or moisture
content, the difference in POCR between test 1 and test 2 did correlate with difference in
temperature (Figure 4-34). Except one point (circled), the increase in temperature

correlated well with POCR increase. Most of the increase in POCR was for temperature
increase up to 12 °C. Additional temperature increase from 12 °C to 27 °C did not give

signiﬁcant additional increase in POCR. The outlier (circled) has an average moisture
content of 0.35, while all other locations had a moisture content of 0.55:0.04. We

believe that this lower moisture content is the reason for the low increase in POCR.
Because there was a large temperature increase (33 °C) at this location even with 0.35

mositure content, activity of microorganisms was not limited. Hence the reason for lower

132

 

increase in POCR maybe related to either smaller heat capacity of waste at lower

moisture content, or to larger void volume of lower moisture content.

 

 

 

 

 

 

 

 

 

N .5

§ . x

'U

g .4.

§ X

a D

8 .3-

E

O

..D

a:

8 .2' x1> .

m Lfﬁ nUIanr

.S x

8 1 D Lift 5

§ - 0 xx APOCR = 0.127371n(AT)—0.0491 0 Lift 3

g R =o.731

D 0.0 X _ - _ >< Lift2
o 10 20 30 40

Difference in temperature between test 1 and test 2

Figure 4-34: POCR Difference between of tests 1 and 2 vs. their temperature difference

4.2.5. Comparison between POCR and OCR from the landfill

Oxygen consumption rate was also determined at the vents at the time of POCR were
measured in the landﬁll. The oxygen concentration in the vents decreased following ﬁrst
order kinetics (Figure 4-35). OCR values are shown as vertical lines for different vents
for tests 1 and 2 in Figures 4-36 and 4-37. The horizontal lines represent POCR

measured inside the landﬁll in terms of mean with 95% conﬁdence interval and range.

133

 

For test 1 vent L32 had a value above the range measured in the landﬁll, while for test 2,
OCR values at the vents were within same range of POCR determined inside the landﬁll.
Moreover, OCR was not measurable at one vent because oxygen concentrations at the
vent were undetectable. These measurements demonstrate that measuring POCR is more
accurate representation for aerobic activity compared to OCR determined at the vents.

From this limited data set, it is not possible to determine the relationship between OCR at

the vents and POCR inside the landﬁll.

 

 

 

 

 

 

 

 

 

 

 

 

 

Si
. : ~ _.__ w_~
i .8
; E ______
C.‘
l s .
t 8 . - - __..— ______
5 °
. OD - _ _ _
>3
X
l o +_
o 1 2 3 4 5 6
J

w Time (hr)

 

Figure 4-35: Oxygen consumption rate at vent L6 (Date: 10/24/03)

134

 

 

 

 

 

 

 

 

 

 

 

 

I LIT i A “-12:35020 Conﬁdence interval for Mean 7

I F O W a 0.2909 #7
I I «r 0.201

, I 0.3593

I I min I I rmx

I I x i / 4 L32 ( Liﬂ 5)

I i / I

I I L6 ’7 -

I I I L32 (hfts 1,2 &3)

I 0 0.1 0.2 0.3 0.4 0.5 0.6
I Rate of oxygen consumption during test 2 (h' 1)

 

 

Figure 4-36: Comparison between POCR and oxygen consumption rate in vents during

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

test 1

I ’ '17.:T—‘17'il ’— "" ..- .H .._~. _- _-_ _..._ _i:__.4__.__.ﬁ,'— _. #——TI
I I --—~><-—~—- 95% Conﬁdence Interval for Mean I I
I

I I . . F 'I

I I 5/ L32 (hits 1, 2 & 3 ) I

I I 3 T 0.1154 I

I . . I» 0.3925 I

I Imin s i m I

I I ' i i ' I

' I x..-” I I
I I ; i L32 ( Lift 5) /' I

‘ | l L6 .

I I_ ’r . . _ ._ . __ 0 I I
I O 0.1 0.2 0.3 0.4 0.5 3
I

Rate of oxygen consumption during test 1 (h' 1) I

Figure 4-37: Comparison between POCR and oxygen consumption rate in vents during
test 2

135

4.3. Evaluation of Methanogensis

The goal of increasing temperature in bioreactor landﬁll is to improve the operating
conditions for methanogenesis. However, intermittent air injection leads to aerobic
conditions within the bioreactor landﬁll. Exposure to air is detrimental to methanogens,
because methanogens are strict anaerobes. Methanogenic activity in bioreactor landﬁll
cell was evaluated in terms of lag time, the rate of increase in percent methane, and
hydrogen accumulation. Lag time was deﬁned as the time during which methane
concentration change was negligible. Methane and hydrogen concentrations were
measured at sampling ports within the landﬁll for almost a year during startup and for
several weeks after stopping air injection. To evaluate the effect of air injection on
methanogenic activity, a comparison of hydrogen concentration, methane concentration

and lag time was made during startup (i.e., before air injection) and alter air injection.

This section is divided into the following: 1) methane generation and hydrogen
accumulation during startup, 2) methane generation and hydrogen accumulation aﬁer air
injection was stopped, 3) factors affecting methanogenesis establishment during the
startup period, 4) factors affecting methanogenesis establishment after air injection was

stopped, and 4) comparison of methanogenic activity between the two periods.

4.3.1. Methane generation and hydrogen accumulation during startup

The startup period was divided into two: lag time period with negligible increase in
methane concentration and a period during which methane concentration was increasing.

Methane concentration remained constant for 137i27 days in lifts 2, 3 and 4, after which

136

it started to increase. For lift 1 methane concentration at day 88 (ﬁrst day of sampling)
was 24.6:5.0 % and continued to increase afterwards. Thus lag time for lift 1 was not
measured, but it is known to be signiﬁcantly lower than other lifts. An example of the
shape of methane concentration-time proﬁle is shown for Figures 4-38 for location G213
in lift 2, and in Figure 4-39 for location G134 in lift 3. The rate of methane percentage
increase was calculated for each individual location in the landﬁll for the active period
regarding methane percent increase. The rate was calculated as the difference between
the initial and ﬁnal methane concentration during the active period divide by the time
elapsed during this change. For location 6213, this rate was calculated as the difference
between the ﬁnal concentration (36.6%) and the background concentration (6.1%),
divided by the time difference (331-222). For location 6134, the active period was the
initial one, and the rate was calculated as the difference between ﬁnal concentration

(60%) and background concentration (24.6%) divided by the time (290-88) days.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

40 __
Q 35 *I' — ——————— V’-
I 8.; 30 L.-- __.__.._ _ﬂ _- __ _- _
t:
.8 25 ‘. 1
a 20 m2 _- -
§ 15 I — —— — I
8 10 I 3___ W a
o 3 o g
;. __ ___ _ 9 _ ____;__ ”bmrfr
1%; 3;: _
' E . - ‘ l T l j
I 0 50 100 150 200 250 300 350

I
I Days ﬁom start of ﬁlling lift 2 (Location 6213)

 

 

r o G213 —LagtimemarkI

 

 

Figure 4-38: Methane concentration as a function of time during startup in a sampling
location in lift 2

137

 

 

 

 

 

 

 

 

 

 

 

 

100 200 300 400 500 I
Days from start of ﬁlling lift 1

Methane concentration (%)
N
o o o o o
v
I
I
I
I
I
I
I

 

 

Figure 4-39: Methane concentration as a function of time during startup in a sampling
location (G134) in lift 1

The average rate of methane percentage increase for the locations in the landﬁll during
the active period was 0.15:0.03% day'l (Figure 4-40). Rate of methane percentage
increase was normally distributed; hence averages would be used to describe this rate
(Figure 4-40). This rate was calculated for 37 locations within the landﬁll (Table 4-22).
Conditions under which the rate was determined such as, initial waste temperature,
temperature of the waste during lag time, temperature of the waste, COD and pH are
shown in Table 4-22. For these locations, the calculated rate of methane percent
increase, hydrogen concentration during and after lag time, lag time, background

concentration and ﬁnal methane concentration are shown in Table 4-23.

138

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Rate of methane generation

 

 

Std. Dev = .09

Mean = .15
N = 37.00

Figure 4-40: Distribution of the rate of increase in methane concentration in Northern
Oaks bioreactor landﬁll during startup

Table 4-22: pH, COD, and temperature at sampling locations during landﬁll startup

 

 

 

 

 

 

 

 

 

 

 

 

Initial waste Temperature
Location Lift number tem erature 0f the waste Temperature COD H
p during the lag of the waste p
(C) time
0112 Liﬁl 35.4 16.5 12.3 38,600 603
G122 Liﬁl 28.7 14.3 12.6 36,000 6.22
6123 Liﬁl 26.8 11.0 10.0 N/A 6.13
0124 Liftl 28.9 N/A N/A 37,800 6.13
0131 Liﬁ 1 24.1 13.5 10.5 N/A
G132 Lift 1 25.7 13.1 12.4 N/A 6.1
0133 Liﬁl 29.0 13.7 12.2 N/A 6.19
0134 Liftl 29.1 14.4 11.2 43,700 6.34
GI35 Liﬁl 28.2 12.3 11.1 39,300 5.96
6142 Liﬁl 22.5 13.8 11.8 N/A N/A

 

139

 

 

Table 4-22: pH, COD, and temperature at sampling locations during landﬁll startup

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

140

(continued)
L'ft Initial waste 1:11;:333 Timlferat“? e
. I o t e waste
Location mber tempzrature during the COD pH
( C) . 0 (°C)
lag time ( C)
0143 Liﬁl 22.1 13.4 12.1 N/A N/A
0144 Liﬁ 1 24.1 N/A N/A N/A N/A
(3145 Liﬁ 1 22.8 13.6 11.6 N/A N/A
G146 L111 1 24.0 12.7 12.2 36,223 5.79
G211 Lift 2 21.0 13.4 13.5 N/A N/A
G213 L111 2 15.5 11.7 12.0 4,822 6.02
G214 Liﬁ 2 13.9 9.4 10.5 N/A N/A
G215 Lift 2 N/A 11.5 11.6 N/A N/A
G221 Liﬁ 2 20.0 14.7 14.0 N/A N/A
G222 L111 2 17.2 11.7 11.9 32,261 6.09
G223 L111 2 23.5 20.9 17.0 34,375 5.91
G225 L111 2 22.5 16.9 17.1 N/A N/A.
0231 Lift 2 14.8 19.8 19.5 26,089 6.23
G232 Liﬁ 2 10.7 10.9 10.7 N/A N/A
G233 Liﬁ 2 12.0 11.0 11.6 N/A N/A
G311 Liﬁ 3 3.9 7.4 8.6 N/A N/A
G312 L111 3 1.7 7.8 8.6 28,076 6.22
G313 L111 3 17.4 14.1 15.5 28,034 5.74
G321 L111 3 0.2 8.1 8.6 N/A N/A
G322 Lift 3 3.9 8.4 8.6 28,372 5.8
G323 L111 3 N/M N/A N/A 15,181 5.61
G411 L111 4 -0.7 3.3 4.8 12,686 6.1
G412 L111 4 2.9 4.8 5.6 21,607 6.08
G413 L111 4 -1.6 0.5 0.7 22,241 6.16
G421 L111 4 -1.0 2.5 3.2 11,502 6.2
G422 L111 4 4.5 4.0 4.5 25,920 6.22
G423 L111 4 7.7 6.9 7.0 27,568 6.06
Total 37 35 34 34 20 23
N/A Not applicable (sample was not obtained)

 

Table 4-23: Hydrogen during and after lag time, initial and ﬁnal methane concentration,
rate of increase in percent methane and lag time at sampling locations during landﬁll

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

startup
Hydrogen Hydrogen Background Final 11:33::
Name co:centrslitlon concentration methane La methane in
uring ag a ter lag time 0 time
time (ppm) (ppm) (AI) (%) percent
methane
G112 N/A 398 27.02 N/A 52.6 0.130
6122 N/A 123 14.11 N/A 61.8 0.150
G123 N/A 4,829 22.85 N/A 62.9 0.130
G124 N/A 162 32.76 N/A 63.8 0.120
G131 N/A 66 43.06 N/A 65.4 0.320
G132 N/A 123 32.07 N/A 66.9 0.130
G133 N/A 325 18.49 N/A 61.1 0.220
G134 N/A 867 18.25 N/A 59.3 0.210
G135 N/A 1,497 32.03 N/A 52.6 0.090
G142 N/A 5,287 18.69 N/A 56.4 0.240
G143 N/A 200 16.35 N/A 55.7 0.200
G144 N/A 587 16.62 N/A 56.7 0.210
G145 N/A 122 19.36 N/A 60.4 0.210
G146 N/A 640 33.3 N/A 65.1 0.170
G21 1 5,451 473 5.02 222 30.9 0.240
G213 8,254 2,618 6.13 222 36.6 0.280
G214 4,846 385 16.67 130 46.7 0.170
G215 9,479 822 15.46 222 47.9 0.381
G221 11,560 7,001 8.15 156 38.3 0.172
6222 383 217 11.99 222 33.7 0.199
G223 N/A 2,883 11.83 40 39.8 0.096
G225 10,948 7,752 3.18 191 42.5 0.339
G231 525 123 3.43 222 23.5 0.184
G232 9,866 12,561 3.11 222 11.2 0.095

 

141

 

Table 4-23: Hydrogen during and after lag time, initial and ﬁnal methane concentration,
rate of increase in percent methane and lag time at sampling locations during landﬁll

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

startup (continued)

H Rate of

Nam 801.33.131.12... 8033353530.. “13:53:31 3.23. 1.53221... “1.1"“

during lag after lag time percent

time (ppm) (ppm) (ppm) (days) (%) methane

(%lday)
G233 5,751 2,276 3.3 102 32.2 0.141
G311 N/A 27,350 0 55 10.5 0.040
G312 N/A 9,267 1.4 68 3.6 0.010
G313 106,414 60,178 0 122 8.1 0.050
G321 29,571 1,097 3.3 157 18.7 0.110
G322 50,055 20,145 0.3 96 16.3 0.080
G323 N/A 816 0.2 41 8.9 0.040
G411 17,167 20,247 0.3 122 7.8 0.050
G412 14,229 18,640 0.7 122 27.7 0.140
G413 17,492 38,161 0.9 122 15.3 0.070
G421 6,622 19,496 0.6 122 21.8 0.1 10
G422 6,486 15,718 0.4 61 13.9 0.090
G423 27,359 52,996 1.3 122 20.5 0.120

37 19 37 37 23 37 37

 

 

 

N/A Not applicable (sample was not obtained)

The average rate of percent methane increase for lifts 1, 2, 3 and 4 were 0.18:0.035,
0.2110062, 0.055i0.037, and 0.097i0.035 % day", respectively (Figure 4-41). This rate
tested to be signiﬁcantly different between the lifts based on ANNOVA test (Table 4-24).
ANNOVA test was used because the rates were normally distributed within each lift (4-
25). The difference between the groups is observed to be between the lower lifts (1-2)

and the upper lifts (3-4). Differences in average rate of increase in percent between lifts

142

1 and 2, and between lifts 3 and 4 were not signiﬁcant. Difference in this rate between

lower (1-2) and upper lifts (3-4) was signiﬁcant based on t-test (Table 4-26). The rate of
increase in percent methane in the lower lifts was 0.193i0.031 % day'1 and for the upper

1111s 007610.025 % day".

 

 

 

 

 

.2:

.1. _I_' __

 

 

 

 

 

 

 

 

 

 

 

 

0.0‘

Rate of increase in methane concentration

 

 

 

N= 14 11 6 6
Liﬁl Lift2 Lift3 Liﬁ4

Figure 4—41: Rate of increase in methane concentration (%/day) for each lift

Table 4-24: ANN OVA test to determine differences between the different lifts in rate of
increase in percent methane

 

 

 

 

Sum of Mean ,
Squares Df Square F 3'8-
Between
Rate of Greg’s 0.1212 3 0.0404
methane , ,
concentration grthm 0.1461 33 0.0044 9.13 0.000152
increase roups
Total 0.2673 36

 

 

 

143

Table 4-25 Normality tests for methane concentration, hydrogen concentration and rate
of increase in percent methane

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Kolmogorov-Smirnov Shapiro-Wilk
L‘ﬁ Statistic df Sig. Statistic df Sig.
number
Liftl 0.121 14 0.200 0.949 14 0.545
Final L1112 0.175 11 0.200 0.929 11 0.399
methane .
concentration L1113 0.205 6 0.200 0.951 6 0.750
L1114 0.213 6 0.200 0.934 6 0.615
Liftl 0.337 14 0.000 0.609 14 0.000
Hydrogen L1112 0.276 11 0.019 0.801 11 0.010
concentration
annlngnme Lift3 0.202 6 0.200 0.862 6 0.196
Lift4 0.356 6 0.017 0.794 6 0.051
Liftl 0.155 14 0.200 0.936 14 0.368
Rate of
methane Lift2 0.178 11 0.200 0.932 11 0.433
concentration L1113 0.223 6 0.200 0.949 6 0.735
increase
Lift4 0.156 6 0.200 0.981 6 0.955
Hydrogen Lift2 0.151 10 0.200 0.909 10 0.276
°°“°e.““a“°“ L1113 0.319 0.108 0.769 5 0.044
during lag
time L1114 0.203 6 0.200 0.912 6 0.452

 

 

Table 4-26: t-test to test the difference between upper and lower lifts with respect to the
rate of increase in methane percent

 

 

95% Confidence
Std. Interval of the
t (if Sig. (2-tailed) .Mean Error Difference
Difference .
Difference
Lower Upper
4.99 35 1.6310'5 0.117 0.023 0.070 0.16

 

 

 

144

To Show the variations within each lift in the methane concentration-time proﬁles, an
example of this proﬁle for all six sampling locations in lift 4 is shown in Figure 4-42.
The lag time for lift 4 was 80 days. The concentration increased after from below 1% to

a concentration range between 12.3 % and 27.7% by day 238 from start of lift 4.

 

 

 

 

 

 

 

 

 

 

 

A 30 I— 2- -~———--- —— —— — ,
_-_w. _ . /3
g 20 I_._. P— _7— "——_—”_—"‘ ”ﬂ” 7 A‘ “ / I
5 15 _I *5 __.__,_____ __.- . - . _w/[q —I
g i / .
8 10 I""_ “u“ — —_ __’ “#1" “’—‘——G
o .
8 s -I--———- —~ » ,/
g 0 1..-».-- - -----» . C .

0 50 100 150 200 250

Days fromﬁllinglift4

 

 

 

Figure 4-42: Methane concentration during start up of bioreactor landﬁll lift 4

Hydrogen concentration during the startup phase of the anaerobic bioreactor landﬁll was
also measured as a second indicator of methanogenic development. Hydrogen
concentrations were much higher than the 100 ppm level known to be thermodynamically
unfavorable to syntrophs (butyrate and propionate utilizers). Hydrogen concentrations
in the waste during and after lag time were not normally distributed (Table 4-25), hence
the median is used to represent the values of hydrogen in the waste. The range of
hydrogen concentrations during lag time for all the locations in the landﬁll was between
383 and 106,414 ppm, with a median of 9,866 ppm. The range of hydrogen

concentration after the lag time was between 66 ppm and 60,178 ppm, with a median of

145

1497 ppm. There was no statistical difference between hydrogen concentration before

and after the lag time, as concluded from Wilcoxon Signed Ranks Test (Table 4—27).

Table 4-27: Wilcoxon Signed Ranks Test to study the difference in hydrogen
concentration between the two periods: before and lag time

H dro en concentration Z Asymp. Sig. (2-tailed) I
l Before and after lag time -0.926‘ 0.355 I

'Based on positive ranks

 

 

To study the variation in the lifts, the median for each lift was calculated. Median
concentration of hydrogen before lag time was 1,549 ppm, 20,145 ppm and 19,872 ppm
for lifts 2, 3 and 4, respectively (Figure 4-43). Lift 1 had lag time equals to zero hence
hydrogen concentration during that time was not applicable. The median of hydrogen
concentration after the lag time was 362 ppm, 2,276 ppm, 14,706 ppm and 19,872 ppm
for lifts 1, 2, 3 and 4, respectively. Concentrations of hydrogen differed signiﬁcantly
between the lifts (Table 4-28). The variability in hydrogen concentrations was big as
indicated by the following ranges observed in each lift. The range of methane
concentration after lag time was from 66 to 5,287 ppm for lift 1, from 123 to 12,561 ppm
for lift 2, and from 816 to 60,178 ppm for lift 3 and from 15,718 to 52,996 ppm. In
addition, hydrogen concentrations during the lag time were different than those after the
lag time, when compared lift by lift (Figure 4-43). Lifts 2 and 3 had signiﬁcantly lower
hydrogen concentrations after the lag time, while lift 4 had signiﬁcantly higher hydrogen

concentrations.

146

Table 4-28: Kruskal Wallis Test to study the difference in hydrogen

concentrations between the different lifts

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Hydrogen Lift Mean Chi- Df Asymp.
concentration number Rank Square Sig. .
Liﬁ2 10 6.50
, , Lift3 18.00
During lag time 10.57 2 0.00507
Lift4 6 11.83
Total 19
Lift2 11 17.45
Lift3 6 27.17
After lag time Lift4 6 32.33 20.40 3 0.00014
Total 37
Liftl 14 11.00
70,000
A 60,0001
E
o.
8: 50,0001
1::
.2
5 40,0001
8 O
0 30,0001
1:
8
5 20,000' -x-
‘10 La time
g 10,0001 3: _ g
>~. .
I 01 _L 9% __ EB; [JDurmg
-10,000 - _ _ _ EjAfter
N= 5 12 12 5 5 5 5
Liftl Lift2 Lift 3 Lift4

Figure 4-43: Hydrogen concentration during and after lag time

147

Hydrogen concentration -time proﬁles for all the locations in lift 4 is Shown in Figure 4-
44. Hydrogen concentrations in lift 4, measured 40 days after the start of lift 4 was at a
median of 7,835 ppm.‘ With time, hydrogen concentrations increased further. The
median H2 concentration after 238 days was 30,308 ppm. The results indicate that even
after 238 days, the landﬁll was severely inhibited. Moreover, the hydrogen concentration
increased over the 238 days, thus indicating the inability of hydrogen utilizing
methanogens to uptake the produced hydrogen. Indeed, these elevated hydrogen

concentrations were accompanied by low methane concentrations (18.59%) (Figure 4-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

41).
i E 120,000 ”_-_ , __
3 100,000
8 80,000
'33
§ 60,000
§ 40,000 I“
O .
g 20,000
.8” °
. E‘ 0 50 100 150 200 250
I Days from ﬁlling lift 4
I

 

Figure 4-44: Hydrogen concentration during start up of bioreactor landﬁll lift 4

Table 4-29: Wilcoxon Signed Ranks Test to test differences between hydrogen
concentration during and after lag time for each lift

 

 

 

 

Z Asymp. Siia-tailed)
lift 2 -2.4973 0.012515
lift 3 -2.0226 0.043114
lift 4 -2.2014 0.027708

 

 

 

148

4.3.2. Methane generation and hydrogen accumulation after stopping air injection

After air injection, percent increase in methane concentration and hydrogen
concentrations were studied only in the locations that were exposed to air. The locations
studied after air injection are shown in Table 4-30, with average waste temperature
during this test, rate of increase in percent methane, and hydrogen concentration. For all
the locations studied, the methane concentration was reduced to a median of 2.5% at the
end of the air injection. Median was used for data that are not normally distributed;
otherwise average was used to represent the center of the data (Table 4—31). Median of
the lag time observed at all studied locations before was 11 days. After lag time, median
of rate of percent increase in methane concentration was 0.255 % day'l. Methane
concentration in the waste reached 40.5:7.6% for lifts l, 2 and 3 after 62 days ﬁ'om
stopping the air injection, and 21.1i8.0% for lift 4 after 40 days. The concentration of
methane was not monitored after the 40 days, due to freezing conditions in the sampling

ports.

To consider the variations in methanogenic development within the waste, variations in
each lift is studied. Representation of the center of the data was made using average
values for normally distributed data, otherwise the median was used (Table 4-32). Rates
of methane generation for lifts l, 2, 3 and 4 were 1.148i0.600, 0.7181016],
0.452i0.215, and 0.494:t0.170 % CH4 per day, respectively (Figure 4-45). Rate of
methane concentration increase was signiﬁcantly different between the lifts based on
ANNOVA test (Table 4-33). Final methane concentrations at day 62 from turning off the

air injection were 60.7i7.3 %, 41.3i7.2 %, and 18.0-336.5 %. Methane concentration for

149

 

lift 4 was 21.1i8.0 % measured on day 40 from turning off the air injection. The
difference in concentration of lifts 1, 2, and 3 was signiﬁcant based on ANNOVA test

(Table 4-33).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

t: 1.8

.2

g 1.61

8 I

g 1.41

8

a) 1.2‘ _—

5 1—1

E. 1.01

d.)

E

a .8‘

o __

58 6I I

at; :—

.2 4‘ —“‘ _—

‘4—1

0 _1_

$3 .21

g 0.0 _ . . _
= 4 11 5 5

Lift 1 Lift2 Lift3 Lift4

Figure 4-45: Rate of increase in methane concentration (%/day) for each lift

150

Table 4-30: Waste temperature, hydrogen concentration, lag time, ﬁnal methane
concentration and rate of increase in percent methane for locations exposed to air

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

H d H d Rate of
ro en ro en . increase
Location temYSSure :uringg 5t" 151g tlir‘i'rge mftgzire in

lag time time percent

(°C) (ppm) (ppm) (days) (%) methane

(%lday)
G141 19.9 N/A 2,372 0 53.5 0.829
G142 19.6 N/A 274 0 56.6 0.825
G143 22.9 N/A 527 0 60.8 1.384
G144 N/A N/A 550 0 64.9 1.554
G21 1 25.3 1,746 963 6 26.5 0.400
G213 19.2 8,103 4,432 11 50.3 0.640
G214 17.5 972 193 19 57.0 0.680
G215 16.4 764 931 28 56.7 1.140
G221 39.8 886 1,949 11 33.1 0.560
G222 22.4 803 320 1 1 36.7 0.420
G223 20.8 1,729 306 28 44.3 0.800
G224 25.5 2,102 1,114 11 51.8 0.830
G225 19.8 6,466 3,325 28 41.7 0.940
G231 16.8 372 233 28 22.1 0.520
G233 14.7 1,658 1,282 28 42.0 0.970
G311 15.7 1,628 988 28 23.9 0.560
G312 14.0 11,008 8,781 28 13.2 0.300
G313 18.9 23,471 34,486 40 21.1 0.670
G321 20.5 1,462 650 28 20.4 0.470
G322 13.7 11,519 13,198 28 12.0 0.260
G41 1 22.6 279 206 6 31.3 0.720
G412 26.4 384 192 6 19.2 0.440
G413 10.0 19,210 61,459 6 15.7 0.370
G421 33.2 1,978 942 6 23.2 0.550
G423 16.8 2,805 2,026 6 16.1 0.390

 

 

 

151

Table 4—31: Normality tests for concentration of methane, ﬁnal methane concentration,
hydrogen concentration and rate of increase in percent methane in the landﬁll as a whole

 

 

 

 

 

 

 

 

 

Kolmogorov-Smirnov Shapiro-Wilk
Statistic df Sig. Statistic df Sig.
BaCkgr°und netham 0.261 27 5.2410'5 0.732 27 1.1210'5
concentration
Fmalmethfme 0.142 27 1.7110'1 0.933 27 0.081
concentration
Hydmg‘?“ “me.“mtw“ 0.350 27 2.8210'9 0.631 27 4.8410'7
during lag trme
Hydmge“ °°“°.ent’a"°” 0.378 27 5.8110’” 0.429 27 3.4710'9
after lag trme
Lagtime 0.250 27 1.4110“1 0.849 27 110103
Rate of increase in
methane concentration 0.134438 25 0.2 0.913981 25 0.037432

 

152

 

Table 4-32: Tests of normality per lift for rate of increase in percent methane, methane

concentration and hydrogen concentration

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Kolmogorov-Smirnov Shapiro-Wilk
”‘1 Statistic df Sig. Statistic df Sig.
number -
Lift] 0.301 4 N/A 0.825 4 0.156
Rate of
increasein Lift2 0.109 11 0.20 0.962 11 0.792
percent Lift3 0.210 5 0.20 0.941 5 0.671
methane .
L1114 0.246 5 0.20 0.883 5 0.323
L1111 0.162 5 0.20 0.969 5 0.866
Final Lift2 0.119 12 0.20 0.960 12 0.779
methane .
concentration L1ﬁ3 0.269 5 0.20 0.883 5 0.321
Lift4 0.217 5 0.20 0.877 5 0.296
Hydrogen L1112 0.344 12 3.27104 0.685 12 6.08104
“minimum Liﬁ3 0.225 5 0.20 0.884 5 0.327
during lag
time Lift4 0.404 5 7.5710'3 0.668 5 4.3210’3
Liftl 0.299 5 0.164 0.849 5 0.191
Hydrogen Lift2 0.249 12 0.039 0.789 12 7.1310'3
concentration .
annlagnme L1113 0.255 5 0.20 0.844 5 0.176
Lift4 0.457 5 9.9610“4 0.576 5 2.9110“1

 

153

 

Table 4-33: ANNOVA table to test the difference in ﬁnal methane concentration in
between lifts 1, 2 and 3 and in rate of increase in percent methane for lifts 1, 2, 3 and 4

 

 

 

 

 

 

 

Sum of Mean .
Squares df Square 813'
Between
Final methane Grows 4567 2 2284
concentration Within 26 3. 54 10-6
(liﬁs 1,2,and 3) Groups 1662 19 87
Total 6229 21
Bé’twee“ 1.321709 3 0.44057
Rate of methane roups
concentration Within 1.201084 21 0.057194 7.7 0.001174
increase Groups
Total 2.522793 24

 

 

 

Figure 4—46 shows the increase in percent methane concentration in lift 2 after air
injection was stopped. The average concentration of methane after stopping the air
injection was 2.1.-1:16 %. The concentration of methane increased at an average rate of
0.72:0.16 % day", reaching an average value of 42.0i6.9 % methane after 62 days.
Change in concentration beyond the 62 day period could not be followed because the
sampling ports were frozen (12/3/03). Methane concentration 4.5 months later (4/20/04)
was 53.8:t7.0, indicating further increase in % methane. During this sampling event,
fewer variations in the methane concentrations were observed. After approximately 6
months from end of air injection test 1, the concentration of methane had less variations
within lift 2 with signiﬁcant methane production. Note that the average concentration of

methane before air injection in lift 2 was 30.8:7.4 %, with a range from 11-50 %.

154

 

 

 

 

 

 

 

 

 

 

 

A70- —— ~——* -—— _-_
c,\" .
:“r‘“‘“ J,,.
.3 50 ”I‘ __f _- L .
§ 40I w«_ —— __.. L3 __.", ...» /
/
g 30 " 725 ~ 7 ‘1
g» 0 W ,_./
0 10 20 30 40 50 60 70
Days ﬁom stopping air injection

 

 

 

Figure 4-46: Methane regeneration after air injection was stopped in lift 2

The concentration of hydrogen was also measured at the end of air injection tests.
Hydrogen concentration during the lag time for all the sampling locations exposed to air
was at a median of 1694 ppm. This hydrogen concentration was variable in the waste
with a range from 279 to 23,471 ppm. After the lag time the concentration of hydrogen
became 964 ppm. The variability in hydrogen concentrations among the sampling
locations was even higher than what was observed during the lag time. The range of
hydrogen concentration after lag time was from 192 ppm to 61,459 ppm. There was no
statistically signiﬁcant difference between the concentration of hydrogen before and after
the lag time, when studied for all the ports at in all the lifts (Table 4-34). Table 4-30

shows for each location the hydrogen concentrations during lag time and after lag time

(Figure 4-47).

155

70,000

 

 

 

 

 

 

 

A 60,000 . *
E
Do
9; 50,000 i
c:
.2
*5 40,000 I
E o
8 30,000 I
c:
8
5 20,000 1
on o
g 10,000 . 3t Lag trme
>1
I 0 I AL 2% D DWI;
-10,000 [:lAfter

 

N= '5 12'12 5'5 5'5
Liftl Lift2 Lift3 Lift4

Figure 4-47: Hydrogen concentration during and after lag time after air injection

Table 4-34: Test statistics for Wilcoxon Signed Ranks Test

H dro en concentration Z Asymp. Sig. (2-tailed) I
I Before and after lag time 4.5422 0.12302 I

'Based on positive ranks

 

 

Median hydrogen concentrations during lag time were 1,315 ppm, 11,008 ppm and 1,978
ppm for lifts 2, 3 and 4, respectively (Figure 4-47). Ranges of hydrogen concentrations
were from 372 ppm to 8,103 ppm for lift 2, from 1462 ppm to 23,471 ppm for lift 2 and
from 279 ppm to 19,210 ppm for lift 4. Lift 1 had zero lag time. Median hydrogen
concentrations after lag time were 550 ppm, 947 ppm, 8781 ppm, and 942 ppm for lifts l,

2, 3, and 4 (Figure 4-47). The ranges for hydrogen concentration after the lag time were

156

ﬁ'om 274 to 2372 ppm for lift 1, from 193 ppm to 4,432 ppm for lift 2, ﬁom 650 ppm to
34,486 ppm for lift 3, and from 192 ppm to 61,459 ppm for lift 4. The hydrogen
concentration data indicated that there is a big variation in hydrogen concentration among
the lifts. The big variation within each lift, and the existence of very low concentrations
in each lift, resulted in a non-signiﬁcant test results between the lifts (Table 4-35).
Hydrogen concentrations during and after lag time were not statistically different
between the lifts, as indicated by Wilcoxon Signed Ranks Test for lifts 3 and 4, however
the concentration after lag time was signiﬁcantly lower than that during lag time (Table
4-36).

Table 4-35: Kruskal Wallis Test to test difference between hydrogen concentrations
among the different lifts

 

 

 

 

 

 

 

 

 

 

Lift N Mean Chi- df' Asymp.
number Rank Square SiL
Lift 2 12 9.83333
Hydrogﬁm °°ncéntrati°n Lia 3 5 16 3.22134 2 0.19975
during lag t1me Lift 4 5 11
Total 22
Lift 1 5 12.4
, Lift 2 12 12.5833
Hydmge“ °°“°.e““a“°“ Lift 3 5 20.2 3.75053 3 0.28969
after lag t1me _
Lift 4 5 12.8
Total 27

 

 

 

Table 4-36: Wilcoxon Signed Ranks Test to study the difference in hydrogen
concentration between before and after lag time

 

Z Asymp. Sig. (Z-tailed)

 

 

 

 

 

lift 2 ~2.118 0.034
lift 3 -0.135 0.893
lift 4 -O.674 0.500

 

157

To further show the type of variations within each lift, an example of hydrogen
concentration-time proﬁle for all the locations in lift 2 is shown in Figure 4-48. The
median of hydrogen concentration after stopping the air injection for lift 2 was 809 ppm,
which increased after 6 days to 2,026 ppm and then decreased to 650 ppm. The
concentration of 650 ppm was observed after 62 days from stopping the air injection.
The marginal increase in hydrogen indicates that hydrogen utilization has improved over
the 62 days operating in anaerobic mode, and short periods of air injection for the
purpose of increasing the temperature can be handled adequately by the methanogens.
The concentration of hydrogen further decrease 4.5 months later (200 days from stopping

the air injection) in lift 2, the median of hydrogen concentration was 278 ppm.

 

16,000 _.__- _
14,000
12,000 . —
10,000 -.

 

 

 
 
  
  

 

 

 

 

 

 

6,000 .
4,000
2,000 .

 

 

 

Hydrogen concentration (ppm)
00
”o
o
O
t
|

Days ﬁ‘om stopping air injection

 

 

 

Figure 4-48: Hydrogen concentration change after air injection was stopped in lift 2

Higher hydrogen accumulation and lower methane concentration at the start up and after
air injection indicated that initial startup of the bioreactor landﬁll was more problematic

compared to the restart period after air injection. The hydrogen concentrations continued

158

to increase during the startup, perhaps because methanogens were yet to establish in a
newly started bioreactor landﬁll. However, during the period following air injection,
hydrogen accumulated to some extent but decreased quickly. The normal methane
concentrations for stable anaerobic digestion (40-60 %) even in the presence of more than
100 ppm hydrogen indicated that the syntrophs may be able to remain active perhaps in a
protected environment that has lower hydrogen concentrations. This indicates that after
62 days from stopping the air injection test 1, methanogenesis was well established with
lower hydrogen concentrations and higher methane concentrations in comparison to those

observed during landﬁll startup.

4.3.3. Factors affecting methanogensis establishment during bioreactor landfill
startup

Waste temperature, pH and COD are some of the most important parameters that affect
anaerobic biodegradation. Methane concentration in the landﬁll did not correlate with
waste temperature of the waste (Figure 4-49, 4-50). Waste temperatures decreased from
26.5 °C to 11.7 °C in lift 1 and from 17.1 to 13.6 °C in lift 2 and remained constant at the
ﬁnal temperature throughout the startup period (Figure 4-49). Methane concentration
increased from 24 % to 50 % in lift 1 and from 8 % to 30 % in lift 2 during the same
period. However, the increase in percent methane in lift 2 had a considerable lag period,
perhaps due to delayed supply of leachate. Waste temperature increased from 3.9 °C to
8.6 °C in lift 3 and from 1.1 °C to 4.6 °C in lift 4, and remained constant at the ﬁnal
temperature (Figure 4—50). Methane concentration in these lifts increased from values

close to 0 % to 19 % in both lifts during the same period. This data supports that

159

 

 

 

methanogenic activity is substantial in bioreactor landﬁlls even at lower temperatures

(e. g., in this case at an average temperature of 10.7i0.7 °C).

The rate of increase in percent methane (computed as difference in % increase divided by
the time between the initial and ﬁnal % methane) did not correlate with waste
temperature (Figure 4-51). The rate of increase in percent methane for lifts 3 and 4 was
less than for lifts l and 2. The time for which lift 2 had constant methane concentration
at approximately 8% was 177 days from the start of lift 2 (Figure 4-49). Lift I remained
open for about 60 days after ﬁlling, while the remaining lifts remained open for only 2
weeks after ﬁlling. The initial temperature of waste (which was kept open until 5 days
after coverage) was different in each lift with temperatures of 26.5, 17.1, 3.9 and 1.1 °C
for lifts 1, 2, 3 and 4, respectively. The exposure to aerobic conditions for varying
amounts of time may be the main reason for the difference in rate of increase in percent
methane. Aerobic biodegradation and hydrolysis at higher temperatures and longer
periods may have lead to the increase in the readily available substrates in the waste for
methanogens. This is supported by a correlation between COD and initial waste
temperature (Figure 4-52). Moreover, methane concentration measured after
approximately 200 days from the start of each lift also correlated with COD (Figure 4-
53). This indicated that hydrolysis of particulate organic matter into soluble COD was
enhanced at a higher rate than hydrolysis observed at locations with lower temperatures.
This process resulted in improved methanogenic activity, measured as increase in %

methane and increase in rate of increase in percent methane.

160

 

 

05
O

 

 

 

 

 

 

 

 

 

 

A ~ 50 28,
e s
2 40 '5
g E
a 3° ‘2’
E o
a 20 3
g 5
. 10 ‘15.
3 2

. . 0

0 100 200 300 400 500

Days
I '7'1—m—‘Tm'im1 . Ten'peraturein 1112‘ ‘ ’

 

. + Metlnne concentration in lift 1 ~-~~><« Methane concentration 'm lift 2

L

 

 

 

 

Figure 4-49: Methane concentration and waste temperature during the startup period for
lifts 1 and 2

 

 

D)
O
1
O\
O

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A 25 - 50 :5
8 J 8
2 20 -» — - 40 IE
'3 e
a 15 a—— ~ - 30 g
E 1 8
g 10 I g
1 § 5 ‘3
( E
7 0
100

 

 

I I 2.-.... Terrperature in; lift 3 -+-— Metlune concentration 'n lift 3

i x Terrperatureinlift4 +Methaneconcennation'nlift4i

 

 

 

 

 

 

Figure 4-50: Methane concentration and waste temperature during the startup period for
lifts 3 and 4

161

Rate of increase in methane concentration

 

 

 

 

.4
D
:1
*-
.3I
1:1
* Cl
9(-
99+
'2‘ a- 0 Liftnumber
1:1 * 1:1
51* OLift4
O * are
11 0 q, a 'Lift3
O
D C’Lift2
0.0 *Liftl
0 10 20 30

Temperature of the waste

Figure 4-51: Rate of increase in methane concentration vs. waste temperature during

 

 

 

 

startup period
50,000
*
40,0001 *
'X-
* -x-
a
A :1
a 30,0001 .
>
3 O O 0 Lift umbe
8 20000 O O n r
U r ' O Liﬁ4
' Lift3
10,0001 8
D Lift2
D
0 _ _ - ' * Lift 1
-10 0 10 20 30 40

Initial waste temperature (C)

Figure 4-52: COD vs. initial waste temperature in the landﬁll

162

7O

 

 

 

 

*-
60' *
8
g 50' * *
b
5
o 40‘
r:
8 Lift number
0 301
5 * 0Lm4
5
a) 201 O .
2 1:1 ' ert 3
O
101 D 00 . O ' c1 ‘3 Lift2
o _ _ °’_ _ *Lmr
0 10,000 20,000 30,000 40,000 50,000
COD

Figure 4-53: Methane concentration after 400 days vs. COD

Rate of increase in % methane did not correlate with pH (Figure 4-54) The pH range in
the bioreactor landﬁll cell was between 5.6 and 6.3 which is on the lower side of the
acceptable pH for methanogens (which is pH 6 to 8.0 [130]). Thus variations in pH were
not expected to be the primary reason for variations in the rate of increase in percent
methane. Nevertheless, all the locations that had hydrogen concentrations above 40,000
ppm had a pH below 6. Except for lift 4, hydrogen concentration decreased with an
increase in methane concentration. The increase in hydrogen in lift 4 may be due to
lower methanogenic activity in comparison to acidogenic activity, i.e., methanogens were
not active enough to remove the hydrogen produced by acidogenes leading to hydrogen

accumulation.

163

 

 

 

 

.5 .3
E Cl
C.‘
o
o
C‘.
8 * ..
o .2' 1:1
5 o
.5 -x-
” «x-
E 0 Lift number
:1 * * x-
.u-a O *
‘1’ O 0 Lift 4
Ki 1 ‘ D
2 . +1- 0 .
a O ’ Lift 3
“5 > ’ O 0 Lift 2
8
g 0.0 ' ' ' * Lift 1
5.6 5.8 6.0 6.2 6.4
pH

Figure 4-54: Rate of increase in methane concentration vs. pH

4.3.4. Factors affecting methanogenesis establishment after air injection was
stopped

Rate of increase in methane concentration and ﬁnal concentration of methane did not
correlate with the waste temperature during the test period (Figures 4-55 and 4-56).
However the rate of increase in percent methane after air injection correlated with the rate
of increase in percent methane during startup at a 2-tailed signiﬁcant level of 0.020
(Pearson correlation coefﬁcient = 0.482). Thus because the spatial trend in rates of
increase did not change it is evident that the inhibitory effects of air injection on
methanogenic activity were temporary. It is possible that methanogens were protected to
an extent from oxygen concentrations in the liquid phase by aggregates. Therefore as

soon as the air injection was stopped the concentration of methane started to increase at a

164

rate that is higher than that obtained before, and was higher for those locations where it

was higher before.

1A

 

Lﬁ

 

Rate of increase in methane concentration
9°

 

 

*
Cl
C]
U
#0 D
O
D E
D Cl
C
D
O
o O D D
D
D
0 1'0 20 30 40 50

Temperature of the waste (C)

Lﬁmmmr
OLift4
’Lift3
DLift2
*Liftl

Figure 4-55: Rate of increase in methane concentration vs. waste temperature after air

injection

165

 

 

 

 

 

70.0
*-

g 60.0« *
'5 DU 9(-
g ..

D
5 50.01 a
O
1::
8 a a”
1, 40.01
g a
a D a
E 30.01 0
_.. D
‘3 . o
i: 20.0. ‘3 n O

o O
10.0 r’
0 10 20 30 40 50

Figure 4-56: Methane concentration after 62 days from stopping air injection for lifts l, 2
and 3 and after 40 days from stopping air inj ection for lift 4 vs. waste temperature

Temperature of the waste

 

1.6

1.4-

1.2-

1.01

.8I

After air injection

.61

.4-

 

*-

 

 

Figure 4-57: Rate of increase in percent methane after air injection vs. before air

Before air injection

injection

166

Liftnumber
C)Lift4
’Liﬂ3
E3Lift2
*Liftl

OLift4
’Liﬁ3
DLiftZ
*Liftl

4.3.5. Comparison of methanogensis establishment between startup and after air
injection

We hypothesized that temperature increase will lead to a higher rate of increase in
percent methane, higher methane concentrations, lower hydrogen concentrations, and
lower lag time. All the parameters used in this study to evaluate anaerobic digestion in
the bioreactor landﬁll indicated a better performance of the bioreactor landﬁll after air
injection was stopped in comparison to the startup period. The average rate of increase
in percent methane at all locations exposed to air in the landﬁll increased signiﬁcantly
from 0.15i0.042 °C day'l to 0.68:0.15°C day“1 (Figure 4-58). This was accompanied by
a decrease in the median of hydrogen concentration from 4,085 ppm to 953 ppm (Figure
4-59 and 4-60). The lag time also decreased from 137 days to 15 days, in response to air
injection. This study demonstrated that air injection is a key to a successful startup of

anaerobic digestion of bioreactor landﬁlls in colder climates.

 

 

 

 

 

 

 

 

 

 

 

 

 

55 2.0

3;

e)

g 1.5. 0*

<6 _.__—

5

0

E

E 1.01

0

2

a

.E -5'

:1:

§ . I 3 —I—

00‘ ._J__.

‘4—

o

8

£3 -.5 _ _
N= 23 23

Startup After air injection

Figure 4-58: Comparison between the rate of methane concentration increase during
startup up and after stopping air injection

167

 

 

 

 

 

 

 

 

 

 

 

a 120,000
C.
3 «x-
; 100,000'
00
-‘—° 80,000:
E”
"O 60,000I
S
.5 o
g 40,0001
2‘3
8 20,000 i
c:
a ,, g
5 01 —l_ m
E

-20,000 . '

= 21 21
Startup Afterairirrjectbn

Figure 4-59: Comparison of the hydrogen concentration during lag time between during
startup up and after stopping air injection

 

 

 

 

 

 

 

 

 

 

 

 

 

8 70,000
ca.
8'
a» 60,0001 0 *
.§
... 0
on 50,0001
..‘9.
l—
a 40,000
CU
*-

5 30,000.
*5
b
5 20,0001
8 -x-
8 10,0001 0
a
a) m
5 °‘
E ~10,000 _

= 24 24

Startup After air injection

Figure 4-60: Comparison of the hydrogen concentration after lag time between during
startup up and after stopping air injection

168

5. CONCLUSIONS

From the experiments conducted in Northern Oaks bioreactor landﬁll, the following
conclusions can be drawn. The conclusions are listed corresponding to each objective for

clarity.

A major part of Objective 1 was to establish and quantify the relationship between the
amount of heat produced (Id) and the amount of oxygen supplied (mole), termed as heat
generation factor. Characterization of performance of the air injection with respect to
temperature and leachate characteristics was also part of Objective 1. The conclusions

drawn from the intermittent air injection experiments are:

0 Heat generation factor was calculated to be 341i84 kJ/mole of oxygen under ﬁeld

conditions.

0 Leachate recirculation can be used as a control mechanism for extreme
temperature increases in bioreactor landﬁlls. This conclusion is drawn because
leachate recirculation, if practiced with leachate temperatures lower than the
waste temperature, leads to immediate heat loss (50% of the total heat loss),

decreasing the waste temperature.

0 Waste temperature increased above the ambient temperatures by 9.6, 13.2, 8.2,

8.4, and 3.6 °C for waste in lifts 1, 2, 3, 4 and 5, respectively. The relationship

169

describing the increase of waste temperature above ambient temperatures was

T -1.1803 T

was-re- ambient

established from the data in this study to be: + 85572

0 Leachate Injection at a rate of approximately 8% (leachate volume to waste
volume) is sufﬁcient to provide a uniform temperature within the waste. This

conclusion was valid at an injected leachate temperature of 125°C and waste

temperature between 4°C and 27 °C.

0 Air injection (21 days at full capacity using 2.5 HP compressor lead to and
increase in waste temperature from 0°C to 30°C for waste with low volumetric
moisture content (0.47:0.03), and from 11°C to 18°C for waste with higher

volumetric moisture content (0.56i0.19) during.

0 In response to air injection (139 days), COD of leachate in lift 1 decreased ﬁ'om

21 g L-l to 2.9 g L-l.

Objective 2 was to determine the POCR and study the effect of moisture and temperature

on POCR. The following conclusions were drawn from the data related to POCR.

0 First order POCR averaged 0.0867i0.0422 h-l after 15 days of air injection. This
increased to 0.2450i0.0843 h-l after 21 days of air injection. The increase in

POCR correlated with increase in temperature for temperature change less than

10°C.

170

Monod equation was not a good ﬁt for POCR, because the range of oxygen
concentrations of 0 to 21% is smaller than the range required to obtain
independent values of Monod coefﬁcients k and Ks. Because it is impossible to
have an oxygen concentration higher than 21% in air, it is not expected that

Monod equation will provide a good ﬁt in such experiments.

Objective 3 was to determine the performance of anaerobic digestion at startup and

following air injection. Performance was evaluated with respect to methanogenesis

establishment, using methane and hydrogen concentrations as indicators for anaerobic

digestion process. The following conclusions were drawn with respect to performance of

anaerobic digestion:

Air injection is key to a successful startup of anaerobic digestion of bioreactor
landﬁlls in colder climates because of its potential to increase the waste
temperature without impacting the methanogenesis. The average rate of increase
in percent methane at all locations exposed to air as a result of air injection
increased signiﬁcantly from 0.15i0.042 °C day"1 to 0.68:1:0.15°C day]. This was
accompanied by a decrease in the median of hydrogen concentration from 4,085
ppm (before air injection) to 953 ppm (after air injection). There was also a
decrease in lag time (i.e., time during which change in % methane is negligible)

also decreased from 137 days to 15 days, in response to air injection.

171

6. ENGINEERING SIGNIFICANCE

Air injection can be used for increasing waste temperature in landﬁlls. Air injection at a
flow rate of 40 to 120 scfrn was practiced, that resulted in a rate of temperature increase
of 0.3 °C/day i 0.3 with a range of 0.1 to 1.6 °C /day. In this study air injection did not
result in ﬁres, even at temperatures above 55 °C, because leachate recirculation was
practiced to reduce waste temperature. In addition, turning off the air injection resulted
in a slow decrease in temperature due to heat conduction within the waste and to heat loss
to the boundaries of the landﬁll. In this study, both leachate recirculation and gas
extraction lines were used to inject air, hence no additional construction was necessary. .
To control temperature, it is recommended that leachate recirculation lines are
constructed in proximity of air injection lines, or place the air injection lines in a manner

that allows use for leachate recirculation.

Heat generation factor can be used to predict temperature increase in the landﬁll cell due
to air injection. In this study, the heat generation factor was computed to be in the range
of 200 to 626 kJ per mole of oxygen, with an average of 341:1:84 lemole of oxygen. The
heat generation factor determined in this study indicates that it is within the same range
reported in literature for waste and other substrates, although variations observed in the
ﬁeld were higher than those obtained in laboratory scale reactors However, heat
generation factor must be determined in other landﬁlls to verify if the values obtained in
this study are signiﬁcantly different from that obtained for pure substrates. The results
from this study support that landﬁlls can be used as a calorimeter to determine tin situ

heat generation factor for solid waste.

172

Point oxygen consumption rate in the landﬁll can be determined in the ﬁeld using ﬁrst
order kinetics. This term was obviously a compromise because of the problems
associated with apportioning a certain amount of solid waste to a measured oxygen
consumption rate. However, it can still be used to assess the aerobic activity of waste in
situ. For waste with higher POCR higher air injection ﬂow rate or higher intensity of air
injection network would be required. POCR found in this study was 0.09 to 0.25 per
hour after 14 and 21 days of air injection. The POCR was affected by moisture content as
well by temperature. This parameter can be used in two dimensional and three
dimensional modeling of oxygen ﬂow within the waste. This may help determine the

distance of travel for the oxygen in the injected air.

Intermittent air injection can be used as means to improve the conditions for anaerobic
digestion process. In this study an air injection up to 25 days at a maximum ﬂow rate of
120 scfrn was use. Subsequent to this air injection event, methanogensis was established
within 60 days as supported by increase in methane concentration to ~ 60% at locations
exposed to air after stopping the air injections. Continuous air injection for more than 25
days was not practiced in this study hence the effect of longer term aeration on the
recovery of methanogens was not studied. These results indicate that intermittent air

injection does not inhibit methanogens for signiﬁcant periods.

Air injection can also be used to improve leachate quality for COD and VOC. In this

study, both COD and VOC concentration decreased after air injection. The mechanism

173

of VOC removal from leachate was not determined.VOC could have either volatilized or

degraded in the process. Further research on this issue need to be conducted.

174

10.

11.

REFERENCES

Pacey, J., D.R. Reinhart, D.L. Hansen, G.T. Townsend, and W.H. Johnson.
Landﬁll Bioreactor-An Innovation in Solid Waste Management. in SWANA
22nd Annual LFG Symposium. 1999.

I SWANA, The Solid Waste Manager's Guide to the Bioreactor Landfill. 2002:

Solid Waste Association of North America.

Wolfe, K., Aerobic bioreactor landﬁll technology: An innovative approach to
municipal solid waste management CE news, 2000: p. 62-67.

USEPA, http://www.epa.gov/epaoswer/non-hw/muncpl/landﬁll/bioreactors.htm.

Warzinski, J ., B.T. Waterrnolen, and DR. Genthe, A superior approach to
recirculation Waste Age, 2000.

Pagano, J .J ., R.J. Scrudato, and GM. Summer, Leachate recirculation at the
Nanticoke Sanitary Landﬁll using a bioreactor trench NYSERDA report 98—6,
1998.

Augenstein, D., R. Yazdani, R. Moore, and K. Dahl, Yolo county controlled
landﬁll demonstration project 1997: p. 43-83.

Moore, R., D. Karina, and R. Yazdani. Hydraulic characteristics of Municipal
Solid Waste: Finding of the Yolo county Bioreactor landﬁll Project. in The
thirteen International Conference on Solid Waste Technology and
Management. 1997. Philadelphia, PA.

Dahl, K. Reuse of shredded waste tires for landﬁll gas collection and leachate

injection systems in Yolo county's landﬁll bioreactor demonstration project. in
Proceedings from SWANA's 21st Annual Landﬁll gas symposium. March
1998. 1998. Yolo county, California.

Townsend, T.G., W.L. Miller, R.A. BISHOP, and 1H. Carter, Combining systems
for Leachate recirculation and landﬁll gas collection Solid waste technologies,
1994: p. 18-24.

Barry, RM. and DC. Demme, Design innovations at the DSWA's latest landﬁll
expansion project WASTECON Proceedings, 1997.

175

12.

13.

14.

15.

l6.

17.

18.

19.

20.

21.

Maier, T.B.e.a., E.S. Steinhauser, N.C. Vasuki, and F.G. Pohland. Integrated
Leachate and Landﬁll Gas Management. in Sardinia 95: Fifth international
Landfill Symposium. 1995. Sardinia, Italy: CISA, Environmental Sanitary
Engineering Center.

Wilson, V.L., W.D. Evans, T.D. Stark, and E. (c/o Repa. An Interim Slope Failure
Involving Leachate Recirculation. in Municipal and Industrial Solid Waste
Disposal Technology Waste Tech 2000. 2000. Orlando, Florida.

Akesson, M. and P. Nilsson, Material dependence of methane production rates in
landﬁlls. Waste Management & Research, 1998. 16(2): p. 108-118.

Lefebvre, X., S. Lanini, and D. Houi, The role of aerobic activity on refuse
temperature rise, 1. Landﬁll experimental study. Waste Management &
Research, 2000. 18(5): p. 444-452.

Green, R.B., W.G. Vogt, and RS. Sullivan. Comparison of emission from
bioreactor and conventional subtitle D landﬁlls. in U. S. EPA workshop on
landfill bioreactors. 2000. Arlington, VA.

Townsend, T.G., W.L. Miller, H]. Lee, and J .F.K. Earle, Acceleration of landﬁll
stabilization using leachate recycle. Journal of Environmental Engineering-
Asce, 1996. 122(4): p. 263-268.

Van Lier, J .B., F. Boersma, M. Debets, and G. Lettinga, High-Rate Thermophilic
Anaerobic Waste-Water Treatment in Compartrnentalized Upﬂow Reactors.
Water Science and Technology, 1994. 30(12): p. 251-261.

Chou, H.-H., J.-S. Huang, and W.-F. Hong, Temperature dependency of granule
characteristics and kinetic behavior in UASB reactors. Journal of Chemical
Technology and Biotechnology, 2004. 79: p. 797-808.

El-Mashad, H.M., G. Zeeman, W.K.P.v. Loon, G.P.A. Bot, and G. Lettinga,
Effect of temperature and temperature ﬂuctuation on thermophilic anaerobic
digestion of cattle manure. Bioresource Technology, 2004. 95: p. 191-201.

Ugwuanyi, J.O., L.M. Harvey, and B. McNeil, Effect of digestion temperature
and pH on treatment efﬁciency and evolution of volatile fatty acids during
thermophilic aerobic digestion of model high strength agricultural waste.
Bioresource Technology, 2005. 96: p. 707-719.

176

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

Kashyap, D.R., K.S. Dadhich, and SK. Shanna, Biomethanation under
psychrophilic conditions: a review. Bioresource Technology, 2003. 87: p.
147153.

Lanini, S., D. Houi, O. Aguilar, and X. Lefebvre, The role of aerobic activity on
refuse temperature rise: 11. experimental and numerical modeling Waste
Management & Research, 2001. 19: p. 58-69.

VanderGheynst, J.S., L.P. Walker, and J.-Y. Parlange, Energy transport in high-
solids aerobic degradation process: mathematical modeling and analysis.
Biotechnology progress, 1997. 13: p. 238-248.

Vollertsen, J., M.D.C. Ahneida, and T. Hvitved-Jacobsen, Effects of temperature
and dissolved oxygen on hydrolysis of sewer solids. Water Research, 1999.
33(14): p. 3119-3126.

Hu, H.-Y., K. Fujie, and K. Urano, Effect of Temperature on the Reaction Rate of
Bacteria Inhabiting the Aerobic Microbial Film for Wastewater Treatment
journal of Fermentation and Bioengineering, 1994. 78(1): p. 100-104.

Kulcu, R. and O. Yaldiz, Determination of aeration rate and kinetics of
composting some agricultural wastes. Bioresource Technology, 2004. 93: p. 49-
57 .

Zinder, S.H., Microbiology of Anaerobic Conversion of Organic Waste to
Methane: Recent Developments. ASM News, 1984. 50(7): p. 294-298.

Shen, CF. and SR. Guiot, Long-Term Impact of Dissolved 02 on the Activity of
Anaerobic Granules. Biotechnology and Bioengineering, 1996. 49: p. 611-620.

Liu, J.-S., I.W. Marison, and U. von Stockar, Microbial Growth by a Net Heat
Up-take: A Calorimetric and Thermodynamic Study on Acetotrophic
Methanogenesis by Methanosarcina bakeri. Biotechnology and Bioengineering,
2001. 75(2): p. 170-180.

Haarstrick, A., DC. Hempel, L. Ostermann, H. Ahrens, and D. Dinkler, Modeling
of the biodegradation of organic matter in municipal landﬁlls. Waste
Management & Research, 2001. 19(320-331).

Morrnile, M.R., K.R. Gurijala, J.A. Robinson, M.J. McInemey, and J .M. Suﬂita,
The importance of Hydrogen in Landﬁll F ermentations. Applied and
Environmental Microbiology, 1996. 62(5): p. 1583-1588.

177

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

Molfetta, AD. and A. Sethi. Heat generation and transport from MSW landﬁll to
groundwater. in Sardinia 2001, Eighth International Waste Management and
Landfill Symposium. 2001. Sardinia, Italy: CISA, Environmental Sanitary
Engineering Center.

Haug, R.T., The practical handbook of compost engineering. 1993: Boca
Raton: Lewis Publishers.

James, A.G., J.A. Watson-Craik, and E. Senior, Use of Three-Stage Laboratory
Model to Determine the Effects of Temperature on Individual Trophic Groups of
Methanogenic Associations Isolated from Anoxic Landﬁll. Journal of Chemical
Technology and Biotechnology, 1996. 67: p. 333-338.

Norstrom, J.M., H.J. Bourque, and RA. Smith. Operation and monitoring of a
bioreactor landﬁll in Sainte Sophie, Quebec. in SWANA 7th Annual Landfill
Symposium. 2002. Louisville, Kentucky: SWANA.

Yazdani, R., J. Kieffer, H. Akau, and D. Augenstein, Monitoring the performance
of anaerobic landﬁll cells with ﬂuids recirculation. 2003, Yolo County, Planning
and Public Works Department- Division of integrated Waste Management:
Woodland. p. 108.

Yoshida, H. and H. Hozumi. Theoretical Study on temperature distribution in
sanitary landﬁll. in proceedings of the second international congress on
Environmental Geotechnics. 1996. Osaka, japan.

Reinhart, D.R., T.G. Townsend, and J. Timmons. Florida bioreactor
demonstration project - Instrumentation. in Waste Tech 2002- Landfill
Conference. 2002.

http://srdata.nist.gov/insulation/Insul query result10.a_sp.
http://hypertextbool_<.com/nhysics/thermal/conduction.

Johnson, W.H. and J. Baker. Operational characteristics and enhanced
bioreduction of municipal waste landﬁll mass by a controlled aerobic process. in
Proceedings from SWANA's 4th Annual Landﬁll Symposium. 1999. Denever,
Colorado: SWANA.

Read, A.D., M. Hudgins, S. Harper, P. Phillips, and J. Morris, The successful
demonstration of aerobic landﬁlling - The potential for a more sustainable solid
waste management approach? Resources Conservation and Recycling, 2001.
32(2): p. 115-146.

178

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

Read, A.D., M. Hudgins, and P. Phillips, Aerobic landﬁll test cells and their
implications for sustainable waste disposal. Geographical Journal, 2001. 167: p.
235-247.

Read, A.D., M. Hudgins, and P. Phillips, Perpetual landﬁlling through aeration of
the waste mass; lessons from test cells in georgia (USA). Waste Management,
2001. 21(7): p. 617-629.

Lee, J.Y., C.H. Lee, and K.K. Lee, Evaluation of air injection and extraction tests

in a landﬁll site in Korea: implications for landﬁll management Environmental
Giology, 2002. 42: p. 945-954.

Wolfe, K. Applied Research: Williamson County Aerated Bioreactor Landﬁll. in
2nd Intercontinental Landfill Research Symposium. 2002. Asheville NC,
USA.

Hater, G.R., G. Hamblin, and D.T. Lane. F ull-Scale Demonstration of a
Sequential Aerobic-anaerobic Landﬁll Bioreactor. in Waste Tech 2000. 2000.
Orlando- Florida.

Smith, M.C., D.K. Gattie, D.D.H. Boothe, and KC. Das, Enhancing aerobic
bioreduction under controlled conditions in a municipal solid waste landﬁll

through the use of air injection and water recirculation. Advances in
Environmental Research, 2000. 3(4): p. 459-471.

Marison, I., A calorimetric investigation of the aerobic cultivation of
Kluyveromyces fragilis on various substrates. Enzyme Microb. Technol., 1987.
9(Jan): p. 32-43.

Cooney, C.L., D.I. Wang, and RI. Mateles, Measurement of heat evolution and
correlation with oxygen consumption during microbial growth Biotechnology
and Bioengineering, 1968. 11(3): p. 269-281.

Biron, B., I.W. Marison, and UV. Stockar, Calorimetric investigation of aerobic
ferrnentations. Biotechnology and Bioengineering, 1987. 30: p. 650-660.

Mou, D.-G. and CL. Cooney, Application of dynamic calorimetry for monitoring
fermentation processes. Biotechnology and Bioengineering, 1976. 18: p. 1371-
1392.

Marison, I.W. and UV. Stockar, The application of novel heat ﬂux calorimeter for
studying growth of Escherichia coli W in aerobic batch culture. Biotechnology
and Bioengineering, 1986. 28: p. 1780-1793.

179

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

Turker, M., Measurement of metabolic heat in a production-scale bioreactor by
continuous and dynamic calorimetry. Chemical Engineering Commission, 2003.
190: p. 573-598.

Luong, J .H.T. and B. Volesky, Determination of the heat of some aerobic
fermentations. The Canadian Journal of Chemical Engineering, 1980. 58: p.
497-504.

Ishikawa, Y. and M. Shoda, Calorimetri analysis of Escherichia coli in continuous
culture. Biotechnology and Bioengineering, 1983. 25: p. 1817-1827.

Bach, P.D., K. Nakasaki, M. Shoda, and H. Kubota, Thermal balance in
composting operations. Ferment. Technology, 1987 . 65(2): p. 199-209.

Adani, F., G. Gigliotti, F. Valentini, and R. Laraia, Respiration index
Determination: A Comparative Study of Different Methods. Compost Science &
Utilization, 2003. 11(2): p. 144-151.

Tremier, A., A.d. Guardia, C. Massiani, E. Paul, and IL Martel, A respirometric
method for characterising the organic composition and biodegradation kinetics
and the temperature inﬂuence on the biodegradation kinetics, for a mixture of

sludge and bulking agent to be co-composted Bioresource Technology, 2005.
96: p. 169-180.

Veeken, A., V.d. Wilde, H. Woelders, and B. Harnelers, Advanced biocoversion
of biowaste for production of a peat substitute and renewable energy.
Bioresource Technology, 2004. 92: p. 121-131.

Gea, T., R. Barrena, A. Artola, and A. Sanchez, Monitoring the biological activity
of the composting process: oxygen uptake rate (OUR), rspirometric index (RI),
and respiratory quotient (RQ). Biotechnology and Bioengineering, 2004. 88(4):
p. 520-527.

Scaglia, B., F. Tarnbone, P.L. Genevini, and F. Adani, Respiration Index
Determination: Dynamic and Static Approaches. Compost Science &
Utilization, 2000. 8(2): p. 90-98.

Adani, F ., P. Lozzi, and P. Genevini, Determination of Biological Stability by
Oxygen Uptake Rate on Municipal Solid Waste and Derived Products. Compost
Science & Utilization, 2001. 9(2): p. 163-178.

Eggen, T. and O. Vethe, Stability Indices for Different Composts. Compost
Science & Utilization, 2001. 9(1): p. 19-26.

180

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

Richard, D. and R. Zimmerman, Respiration Rate - Reheating Potential: A
Comparison of Measures of Compost Stability. Compost Science & Utilization,
1995. 3(2): p. 74-79.

Koening, A. and Q.H. Bari, Application of Self-Heating Test for Indirect
Estimation of Respirometric Activity of Compost: Theory and Practice. Compost
Science & Utilization, 2000. 8(2): p. 99-107.

Adani, F., P. Genevini, and F . Tambone, A New Index of Organic Matter
Stability. Compost Science & Utilization, 1995. 3(2): p. 25-37.

Iannotti, D.A., M.E. Grebus, B.L. Toth, L.V. Madden, and HA]. Hoitink,
Oxygen Respirometry to Assess Stability and Maturity of Composted Municipal
Solid Waste. Journal of Environmental Quality, 1994. 23: p. 1177-1183.

Lasaridi, KB. and BI. Stentiford, A simple respirometeric technique for assessing
compost stability. Water Research, 1998. 32(12): p. 3717-3723.

Randolph, T.W., I.W. Marison, D.E. Martens, and UV. Stockar, Calorimetric
Control of Fed-Batch Fermentations. Biotechnology and Bioengineering, 1990.
36: p. 678-684.

Liang, C., K.C. Das, and R.W. McClendon, The inﬂuence of temperature and
moisture contents regimes on the aerobic microbial activity of a biosolids
composting blend Bioresource Technology, 2003. 86: p. 131-137.

Adani, F., L. Scatigna, and P. Genevini, Biostabilization of mechanically
separated municipal solid waste fraction Waste Management & Research,
2000. 18(5): p. 471-477.

Finger, S.M., R.T. Hatch, and TM. Regan, Aerobic Microbial grth in
semisolid matrices: Heat and mass transfer limitations. Biotechnology and
Bioengineering, 1976. 18: p. 1193-1218.

Liwarska-Bizukojc, E. and S. Ledakowicz, Stoichiometry of the aerobic
biodegradation of the organic fraction of municipal solid waste (MSW).
Biodegradation, 2003. 14(1): p. 51-56.

Nielsen, H. and L. Berthelsen, A Model for temperature Dependency of
Therrnophilic Composting Process Rate. Compost Science & Utilization, 2002.
10(3): p. 249-257.

181

77.

78.

79.

80.

81.

82.

83.

84.

85.

86.

Weppen, P., Determining Compost Maturity: Evaluation of Analytical Properties.
Compost Science & Utilization, 2002. 10(1): p. 6-15.

Beck-Friis, B., S. Smars, H. Jonsson, Y. Eklind, and H. Kirchmann, Composting
of source-separated household organics at different oxygen levels: Gaining an

understanding of emission dynamics. Compost Science & Utilization, 2003.
11(1): p. 41-50.

Veeken, A., V.d. Wilde, and B. Hamelers, Passively Aerated Coomposting of
Straw-Rich Pig Manure: Effect of compost bed porosity. Compost Science &
Utilization, 2002. 10(2): p. 114-128.

Oppenheimer, J .R., A.G. Martin, and LP. Walker, Measurements of air-ﬁlled
porosity in unsaturated organic matrices using a pycnometer. Bioresource
Technology, 1997. 59: p. 241-247.

Richard, T.L., B. Hamelers, A. Veeken, and T. Silva, Moisture Relationships in
Composting Processes. Compost Science & Utilization, 2002. 10(4): p. 286-302.

Ahring, B.K., A.A. Ibrahim, and Z. Mladenovska, Effect of temperature increase
from 55 to 65 degrees C on performance and microbial population dynamics of an
anaerobic reactor treating cattle manure. Water Research, 2001. 35(10): p. 2446-
2452.

Yadvika, Santosh, T.R. Sreekrishnan, S. Kohli, and V. Rana, Enhancement of
biogas production from solid substrates using different techniques - a review.
Bioresource Technology, 2004. 95: p. 1-10.

de Cortazar, A.L.G., J.H. Lantaron, O.M. Fernandez, I.T. Monzon, and M.F.
Lamia, Modelling for environmental assessment of municipal solid waste landﬁlls
(Part 2: Biodegradation). Waste Management & Research, 2002. 20: p. 514-
528.

Veeken, A., S. Kalyuzhnyi, H. Scharff, and B. Hamelers, Effect of pH and VFA
on hydrolysis of organic solid waste. Journal of Environmental Engineering-
Asce, 2000. 126(12): p. 1076-1081.

Masse, DJ. and R.L. Droste, Comprehensive model of anaerobic digestion of
swine manure slurry in a sequencing batch reactor. Water Research, 2000.
34(12): p. 3087-3106.

182

87.

88.

89.

90.

91.

92.

93.

94.

95.

96.

F ukuzaki, S., N. Nishio, and S. Nagai, Kinetics of the Methanogenic Fermentation
of Acetate. Applied and Environmental Microbiology, 1990. 56(10): p. 3158-
3163.

Moletta, R., D. Verrier, and G. Albagnac, Dynamic Modeling of Anaerobic
Digestion. Water Research, 1986. 20(4): p. 427-434.

Vavilin, V.A. and L.Y. Lokshina, Modeling of Volatile Fatty Acids Degradation
Kinetics and Evaluation of Microorganism Activity. Bioresource Technology,
1996. 57: p. 69-80.

Barlaz, M.A., D.M. Schaefer, and R.K. Ham, Bacterial population development
and chemical characteristics of refuse decomposition in a simulted sanitary
landﬁll. Applied and Environmental Microbiology, 1989. 55: p. 55-65.

Wang, Y.S., W.S. Odle, W.E. Eleazer, and MA. Barlaz, Methane potential of
food waste and anaerobic toxicity of leachate produced during food waste
decomposition. Waste Management & Research, 1997. 15(2): p. 149-167.

Pohland, F.G. and J .C. Kim, Microbially mediated attenuation potential of landﬁll
bioreactor systems. Water Science and Technology, 2000. 41(3): p. 247-254.

Cecchi, F., P. Pavan, A. Musacco, J. Mataalvarez, C. Sans, and D. Defaveri,
Monitoring a Fast Thermophilic Re-Start-up of a Digester Treating the Organic
Fraction of Municipal Solid-Waste. Environmental Technology, 1993. 14(6): p.
517-530.

Vieitez, E.R., J. Mosquera, and S. Ghosh, Kinetics of accelerated solid-state
fermentation of organic- rich municipal solid waste. Water Science and
Technology, 2000. 41(3): p. 231-238.

Grifﬁn, M.E., K.D. McMahon, R.I. Mackie, and L. Raskin, Methanogenic
Population Dynamics during Start-Up of Anarobic Digesters Treating Municipal
Solid Waste and Biosolids. Biotechnology and Bioengineering, 1998. 57(3): p.
342-355.

Masse, D.I., L. Masse, A. Verville, and S. Bilodeau, The Start-up of anaerobic
sequencing batch reactors at 20C and 25C for the treatment of slaughterhouse
wastewater. Journal of Chemical Technology and Biotechnology, 2001. 76: p.
393-400.

183

97.

98.

99.

100.

101.

102.

103.

104.

105.

106.

Nachaiyasit, S. and DC. Stuckey, Effect of Low Temperatures on the
Performance of an Anaerobic Bafﬂed Reactor (ABR). Journal of Chemical
Technology and Biotechnology, 1997. 69: p. 276-284.

Liu, J ., G. Olsson, and B. Matiasson, Monitoring and Control of an Anaerobic
Upﬂow Fixed-Bed Reactor for Hi gh-Loading-Rate Operation and Rejection of
Disturbances. Biotechnology and Bioengineering, 2004. 87(1): p. 43-53.

Vavilin, V.A., L.Y. Lokshina, S.V. Rytov, and OR. Kotsyurbenko, Description of
two-step kinetics in methane formation during psychrophilic H2/C02 and
mesophilic glucose conversions. Bioresource Technology, 2000. 71: p. 195-209.

Kettunen, RH. and J .A. Rintala, The effect of low temperature (5-29 degrees C)
and adaptation on the methanogenic activity of biomass. Applied Microbiology
and Biotechnology, 1997. 48(4): p. 570-576.

Muralidharan, V., K.D. Rinker, I.S. Hirsh, E.J. Bouwer, and RM. Kelly,
Hydrogen Transfer Between Methanogens and F errnentive Heterotrophs in

Hyperthermophilic Cultures. Biotechnology and Bioengineering, 1997. 56(3): p.
268-278.

Schmidt, IE. and BK. Ahring, Review: Granular Sludge Formation in Upﬂow
Anaerobic Sludge Blanket (UASB) Reactors. Biotechnology and
Bioengineering, 1996. 49: p. 229-246.

Dinopoulour, G., R.M. Sterritt, and J.N. Lester, Anaerobic Acidogenesis of a
Complex Wastewater: 11. Kinetics of Growth, Inhibition, and Product Formation.
Biotechnology and Bioengineering, 1987. 31: p. 969-978.

Cord-Ruwisch, R., T.I. Mercz, C.-Y. Hoh, and GE. Strong, Dissolved Hydrogen
Concentration as an On-Line Control Parameter for the Automated Operation and
Optimization of Anaerobic Digesters. Biotechnology and Bioengineering, 1997.

56(6): p. 627-634.

Fukuzaki, S., N. Nishio, M. Shobayashi, and S. Nagai, Inhibition of the
Fermentation of Propionate to Methane by Hydrogen, Acetate, and Propionate.
Applied and Environmental Microbiology, 1990. 56(3): p. 719-723.

Hill, D.T. and SA. Cobb, Modeling Predictive Indicators in Animal Waste
Methanogenesis. Transactions of the Asae, 1993. 36(3): p. 879-885.

184

107.

108.

109.

110.

111.

112.

113.

114.

115.

116.

117.

Ahring, B.K., P. Westermann, and RA. Mah, Hydrogen inhibition of acetate
metabolism and kinetics of hydrogen consumption by Methanosarcina
therrnophila TM-l. Archives of Microbiology, 1991. 157: p. 38-43.

Angelidaki, I., L. Ellegaard, and BK. Ahring, A Comprehensive Model of
Anaerobic Bioconversion of Complex Substrates to Biogas. Biotechnology and
Bioengineering, 1999. 63(3): p. 363-372.

Zhao, X., S. Maher, R. Musleh, M. Khire, T. Voice, and S. Hashsham. Bioreactor
landﬁll research and demonstration project at the Northern Oaks Landﬁll, MI. in
Waste Tech. 2003. New Orleans, Louisiana: NSWMA WASTEC.

Lide, D.R., ed. CRC handbook of chemistry and physics. 72 ed. 1991/ 1992,
CRC Press, Inc.

Montgomery, J .M., Water Treatment Principles and Design. 1985, New York:
Wiley-Interscience.

Snoeyink, V.L. and D. Jenkins, Water Chemistry. 1980, New York: John Wiley
& Sons.

Sponza, D.T. and M. Isik, Decolorization and inhibition kinetic of Direct Black
39 azo dye with granulated anaerobic sludge. Enzyme and Microbial
Technology, 2004. 34: p. 147-158.

Magbanua, B.S.J., Y.-T. LU, and C.P.L.G. JR, A technique for obtaining
representative biokinetic parameter values from replicate sets of parameter
estimates. Water Research, 1998. 32(3): p. 849-855.

Ribes, J ., K. Keesman, and H. Spanjers, Modelling anaerobic biomass grth
kinetics with a substrate threshold concentration. Water research, 2004. 38: p.
45 02-45 10.

Goudar, CT. and T.G. Ellis, Explicit oxygen concentration expression for
estimating extant biodegradation kinetics ﬁom respirometric experiments.
Biotechnology and Bioengineering, 2001. 75(1).

Spear, J .R., L.A. Figueroa, and 3D. Honeyman, Modeling the removal of
uranium U(VI) from aqueous solutions in the presence of sulfate reducing
bacteria environmental Science and Technology, 1999. 33‘(15): p. 2667-2675.

185

118.

119.

120.

121.

122.

123.

124.

Knightes, CD. and CA. Peters, Statistical Analysis of Nonlinear parameter

estimation for Monod biodegradtion kinetics using bivariate data. Biotechnology
and Bioengineering, 2000. 69(2).

Smith, L.H., P.L. McCarty, and PK. Kitanidis, Spreadsheet method for evaluation
of biochemical reaction rate coefﬁcients and their uncertainties by weighted

nonlinear least-squares analysis of the integrated monod equation Applied and
Environmental Microbiology, 1998. 64(6): p. 2044-2050.

Liu, L., Z. Wang, J. Yao, X. Sun, and W. Cai, Investigation on the properties and
kinetics of glucose-fed aerobic granular sludge. Enzyme and microbial
technology, 2005. 36: p. 307-313.

Fu, W. and AP. Mathews, Lactic acid production from lactose by Lactobacillus
plantarum: kinetic model and effects of pH, substrate, and oxygen Biochemical
Engineering Journal, 1999. 3: p. 163-170.

Alonso, C., X. Zhu, M.T. Suidan, B.R. Kim, and B.J. Kim, Parameter estimation
in bioﬁlter systems. environmental Science and Technology, 2000. 34: p. 2318-
2323.

Tien, TH. and A. Katayama, Biodegradation kinetics of volatile hydrophobic
organic compounds in cultures with variable fractional volumes. Biotechnology
and Bioengineering, 2004. 85(6).

Grannemann, N.G., R.J. Hunt, J .R. Nicholas, T.E. Reilly, and TC. Winter, The
Importance of Ground Water in the

Great Lakes Region. 2000, US. GEOLOGICAL SURVEY: Water-Resources

125.

126.

127.

128.

Investigations: Lansing, Michigan. p. 19.

Gass, T.E., Geothermal Heat Pumps. Geothermal Resources Council Bulletin,
1982. 11(11): p. 3-8.

Aichele, S.S., Ground-Water Quality Atlas of Oakland County, Michigan. 2000,
US Geological Survey Water-Resources Investigations.

Todd, D.K., Groundwater Hydrology. 1980, New York: John Wiley and Sons.
516.

Dubrunfaut, M., CR. Hebd. Seances Acad. Sci., 1856. 42: p. 945.

186

129. Association, A.W.W., Water quality and treatment : a handbook of
community water supplies, ed. R.D. Letterman. 1999, New York: McGraw-Hill.

130. Speece, R.E., Anaerobic biotechnology for industrial wastewaters. 1996,
Nashville: Archae Press.

187