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 'PDATE DUE
W}; ‘07. “”120.
SEP4§120U7 5515558

 

 

 

 

 

 

 

 

 

 

 

 

 

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

 

LEACHATE RECIRCULATION IN BIOREACTOR
LANDFILLS: FIELD-SCALE TESTING AND MODELING

By

Mazen Mohamad Haydar

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

LEACHATE RECIRCULATION IN BIOREACTOR
LANDFILLS: FIELD-SCALE TESTING AND MODELING

By

Mazen Mohamad Haydar

Leachate recirculation (LR) in municipal solid waste (MSW) landﬁlls operated as
bioreactors offers signiﬁcant economical and environmental beneﬁts. Leachate
recirculation systems (LRSs) consisting of horizontal trenches and vertical wells are
conventional methods. However, need for excavation, relatively high installation
costs and non-uniform distribution of recirculated leachate are the key disadvantages
of these conventional methods. In this dissertation, a new design concept called
permeable blankets (PBs) was evaluated numerically and tested in the ﬁeld using
instrumented test sections. This design concept consists of placing a relatively thin
and high hydraulic conductivity material on a relatively ﬂat or inclined waste surface
in a landﬁll. A perforated pipe is embedded in the PB in the direction parallel to the
shorter or longer plan view dimension of the PB where leachate is injected under a
positive pressure. PBs require no excavation, can substitute multiple horizontal
trenches or vertical wells and achieve a relatively uniform wetting of waste. PBs
made up of granular materials can also provide an ideal platform to embed sensors for
monitoring the pressure, temperature, and migration of injected leachate. Field-scale
testing of PBs was conducted at an active MSW landﬁll located in Jackson Michigan
where three 60-m-wide by IO-m-long PBs made up of crushed recycled glass,
shredded tires and a geocomposite drainage layer (GDL) were constructed. Total

about 50 sensors were embedded in these blankets to measure moisture content, pore-

water pressure, temperature, and vertical stress. Leachate was injected at ﬂow rates
ranging from 0.9 m3/hr to 3.6 m3/hr per unit meter length of the injection pipe. The
data collected from the sensors during the period from September 2003 to May 2005
indicated that the injected leachate traveled across the entire width of the blankets in
time periods ranging from a few minutes to a few hours depending upon the injection
rate. The key conclusion of this study is that permeable blankets can be used as an
new LRS in MSW landﬁlls operated as bioreactors. The key parameters that inﬂuence
pore water pressure in the blanket are: leachate injection rate; and (2) the hydraulic
properties and water contents of the PB and waste. The pore water pressure in PBs

does not exceed the leachate injection pressure head as long as perched water tables

do not exist in the landﬁll.

DEDICATION

I thank God and my family for the completion of this work.

I owe everything to the continuous love and prayer of my father (Mohamad), my
mother (Rasmie), my sister (Abir) and my brother (Rabih).

I thank all those who supported me, prayed for me and wished me well and success
(family members, dear friends, and students).
My prayers go to the soul of my grandparents and to the soul of my young cousin
Zeina.
All the gratitude goes to my teachers and my colleagues in school, in the
Beirut Arab University and in the
Michigan State University.

My deepest love goes to my family and to my country, Lebanon.

iv

ACKNOWLEDGMENTS

I would like to extend my sincere gratitude and appreciation to my advisor,

Dr. Milind V. Khire, for his mentoring, continuous support and guidance during this

work. I feel fortunate that I was his ﬁrst masters and Ph.D. student. His high

expectation of me encouraged and motivated me to excel on the personnel and
academic level.

My special thanks to the members of my Ph.D. committee: Dr. Shuguang Li,

Dr. Xuede Qian, Dr. Roger Wallace, and Dr. Gary Weissman for their valuable
comments and suggestions while reviewing this report.

My sincere appreciation goes to the Department of Civil and Environmental
Engineering, Michigan State University, for partially funding my graduate tenure as a
teaching assistant for Soil Mechanics (CE 312) lab and for granting me graduate
fellowships.

The ﬁeld-scale testing was jointly funded by Waste Management Inc.,
Environmental Research & Education Foundation (EREF), and the National Science
foundation (Grant No. CMS — 0510091). I sincerely appreciate the support from the
sponsors. I would also like to express sincere appreciation to Ronald “Hog”
F eldkamp, Chester Stanley, and Paul Mazanec for coordination of numerous leachate
recirculation trials at the McGill Landﬁll. I also appreciate the assistance provided by

Dr. Xianda Zhao during the ﬁeld instrumentation of the blanket.

TABLE OF CONTENTS

LIST OF TABLES .................................................................................................... x
LIST OF FIGURES .................................................................................................. xi
KEY TO ABBREVIATIONS ................................................................................. xvi
KEY TO SYMBOLS ............................................................................................. xvii
INTRODUCTION ..................................................................................................... 1

BACKGROUND ................................................................................................... 1
CONVENTIONAL LEACHATE RECIRCULATION METHODS ....................... 2
NEW LEACHATE RECIRCULATION METHOD ............................................... 3
OBJECTIVES AND METHODOLOGY ............................................................... 7
DISSERTATION ORGANIZATION .................................................................... 8

PAPER NO. 1: LEACHATE RECIRCULATION USING HORIZONTAL

TRENCHES IN BIOREACTOR LANDFILLS ....................................................... 10
ABSTRACT ........................................................................................................ 10
INTRODUCTION ............................................................................................... l 1
NUMERICAL MODELING ................................................................................ 13
HYDRUS-ZD Computer Model ....................................................................... 13
Conceptual Model and Assumptions ................................................................ 14
Boundary Conditions and Mass Balance .......................................................... 16
Hydraulic Properties ........................................................................................ 17
Dimensions ...................................................................................................... 20
Modeling Results ................................................................................................. 21
Effect of Leachate Injection Pressure Head ...................................................... 22
Effect of Hydraulic Conductivity of Trench Backﬁll ........................................ 25
Effect of Horizontal Trench Geometry and Size ............................................... 26
Effect of Horizontal Trench Spacing ................................................................ 26
Effect of Vertical Spacing ................................................................................ 33
Effect of Geometric F ormation of HTS ............................................................. 36
MODEL VALIDATION ..................................................................................... 40
McCreanor (1998) ........................................................................................... 40
Bachus et a1. (2002) ......................................................................................... 41
Doran ( 1999) ................................................................................................... 43
SUMMARY AND PRACTICAL IMPLICATIONS ............................................ 44
ACKNOWLEDGEMENT ................................................................................... 45
REFERENCES .................................................................................................... 47

PAPER NO. 2: EVALUATION OF HETEROGENEITY AND ANISOTROPY OF
WASTE PROPERTIES ON LEACHATE RECIRCULATION IN BIOREACTOR

LANDFILLS ........................................................................................................... 50
ABSTRACT ........................................................................................................ 50
INTRODUCTION ............................................................................................... 50

vi

HYDRAULIC PROPERTIES OF MSW .............................................................. 52

HORIZONTAL TRENCH LEACHATE RECIRCULATION SYSTEM ............. 54
METHODOLOGY .............................................................................................. 55
Conceptual Model ............................................................................................ 56
Numerical Model ............................................................................................. 56
Monte Carlo Analysis ...................................................................................... 60
RESULTS ........................................................................................................... 64
Homogeneous and Isotropic Properties ............................................................ 64
Heterogeneous and Anisotropic Properties ....................................................... 66
Leachate Flow Patterns .................................................................................... 68
SUMMARY AND CONCLUSIONS ................................................................... 72
ACKNOWLEDGEMENTS ................................................................................. 73
REFERENCES .................................................................................................... 74
PAPER NO. 3: NUMERICAL EVALUATION OF PERMEABLE BLANKETS FOR
LEACHATE RECIRCULATION IN MSW LANDF ILLS ...................................... 77
ABSTRACT ........................................................................................................ 77
INTRODUCTION ............................................................................................... 77
NUMERICAL MODELING ................................................................................ 79
HYDRUS-2D Computer Model ....................................................................... 79
Conceptual Model and HYDRUS-ZD Input ..................................................... 82
Boundary Conditions and Mass Balance ............................... I ........................... 82
Hydraulic Properties ........................................................................................ 83
Geometry and Dimensions ............................................................................... 83
SIMULATION RESULTS .................................................................................. 84
Horizontal Trench ............................................................................................ 84
Permeable Blanket ........................................................................................... 84
Effect of Injection I-Iead ............................................................................... 86
Effect of Width of Permeable Blanket .......................................................... 86
Effect of Hydraulic Conductivity of Permeable Blanket ............................... 88
Permeable Blanket and Horizontal Trench Equivalency ............................... 88
Effect of Settlement of Permeable Blanket ................................................... 90
SUMMARY AND CONCULSIONS ................................................................... 90
ACKNOWLEDGEMENTS ................................................................................. 93
REFERENCES .................................................................................................... 94
PAPER NO. 4: Leachate Recirculation Using Horizontal Permeable Blankets in
BIOREACTOR Landﬁlls ........................................................................................ 95
ABSTRACT ........................................................................................................ 95
BACKGROUND ................................................................................................. 95
LEACAHTE RECIRCULATION USING PERMEABLE BLANKETS .............. 97
Design and Construction Aspects ..................................................................... 97
Advantages of Permeable Blankets .................................................................. 99
LEAHCATE FLOW IN WASTE ....................................................................... 100
NUMERICAL MODELING OF PERMEABLE BLANKETS ........................... 101
HYDRUS-ZD Computer Model ..................................................................... 101
Conceptual Model .......................................................................................... 102
Boundary Conditions ..................................................................................... 104

vii

MODELING RESULTS .................................................................................... 106

Hydraulic Properties of Waste ....................................................................... 109
Saturated Hydraulic Conductivity of MS W ................................................. 1 10
Unsaturated Hydraulic Properties of Waste ............................................... 112
Saturated Hydraulic Conductivity of Permeable Blanket ............................ l 12
Hysteresis in Hydraulic Conductivity of Permeable Blanket ....................... 1 15

Geometry of Permeable Blanket .................................................................... 115
Permeable Blanket Depth .......................................................................... 1 15
Permeable Blanket Width and Vertical Spacing ......................................... 117
Settlement of Permeable Blanket ................................................................ 1 18

Leachate Dosing Frequency ........................................................................... 118

Degrees of Saturation of Waste and Permeable Blanket ................................. 122
Initial Degree of Saturation of Waste ......................................................... 124
Initial Degree of Saturation of Permeable Blanket ..................................... 124

FIELD-SCALE TESTING OF PERMEABLE BLANKET ................................ 127

Field Installation of PB .................................................................................. 127

Automated Monitoring System ...................................................................... 128

Field Data versus Simulated Results .............................................................. 128

SUMMARY AND PRACTICAL IMPLICATIONS .......................................... 132
ACKNOWLEDGEMENTS ............................................................................... 134
REFERENCES .................................................................................................. 135
PAPER NO. 5: GEOTECHNICAL SENSING SYSTEM TO MONITOR INJECTED
LIQUIDS IN LANDFILLS ................................................................................... 138
INTRODUCTION ............................................................................................. 1 3 8

Leachate Recirculation Using Permeable Blankets ......................................... 140

Common Sensors Used in Landﬁlls ............................................................... 140
Moisture Content Sensors .......................................................................... 142

AUTOMATED SENSING SYSTEM ................................................................ 143
Electrical Impedance Moisture Content Sensors ............................................. 145
FIELD INSTRUMENTATION ......................................................................... 149
Monitoring System ........................................................................................ 150
Laboratory Testing and Calibration ................................................................ 153
RESULTS ......................................................................................................... 155

Detection of Leachate Migration in Permeable Blankets ................................ 155

Response of Pressure Transducers and Thermistosr ....................................... 158

Limitations of Sensor System to Detect Leachate Migration .......................... 160

Drainage of Permeable Blanket ...................................................................... 162

Effect of Diurnal Air Temperature on Measurements ..................................... 163

SUMMARY AND CONCLUSIONS ................................................................. 165
ACKNOWLEDGMENTS ................................................................................. 166
REFERENCES .................................................................................................. 167
PAPER NO. 6: LEACHATE RECICRCULATION IN BIOREACTOR LAN DFILLS
USING GEOCOMPOSITE DRAINAGE MATERIAL ......................................... 169
ABSTRACT ...................................................................................................... 169
BACKGROUND ............................................................................................... 1 70
Leachate Recirculation Using Permeable Blankets ......................................... 171

viii

OBJECTIVES ................................................................................................... 172

GEOCOMPOSITE DRAINAGE LAYER ......................................................... 173
NUMERICAL MODELING OF PERMEABLE BLANKETS ........................... 174
FIELD TEST SECTION AND METHODOLOGY ............................................ 177
Test Section Layout ....................................................................................... 177
Embedded Sensors ......................................................................................... 179
Monitoring System ........................................................................................ 182
RESULTS ......................................................................................................... 183
Leachate Recirculation Trials ......................................................................... 183
Veriﬁcation of Response of Moisture Content Sensors ................................... 183
Veriﬁcation of Response of Temperature and Pressure Transducer Sensors 185
Flow of Leachate through the GDL Blanket ................................................... 187
Effect of Injection Rate on Leachate Travel ................................................... 189
Correlation of Modeling and Field Results ..................................................... 191
Monitoring of Pressure Head and Temperature in the GDL Blanket ............... 195
SUMMARY AND CONCLUSIONS ................................................................. 199
ACKNOWLEGEMENTS .................................................................................. 200
REFERENCES .................................................................................................. 202
SUMMARY AND PRACTICAL IMPLICATIONS .............................................. 204
HORIZONTAL TRENCHES ............................................................................ 204
PERMEABLE BLANKETS .................................................... . ......................... 205

ix

LIST OF TABLES

PAPER NO. 1: LEACHATE RECIRCULATION USING HORIZONTAL
TRENCHES IN BIOREACTOR LANDFILLS

Table 1. Saturated and unsaturated hydraulic parameters input to HYDRUS-ZD
to simulate leachate recirculation system consisting of horizontal
trenches. ............................................................................................ 18

Table 2. Effect of vertical spacing between HTs on the hydraulic efﬁciency of
LRS. .................................................................................................. 35

Table 3. Comparison of simulation results from SUTRA-2D and HYDRUS-ZD.
................................................................................................................... 42

PAPER NO. 2: EVALUATION OF HETEROGENEITY AND ANISOTROPY OF

WASTE PROPERTIES ON LEACHATE RECIRCULATION IN
BIOREACTOR LANDFILLS

Table 1. Hydraulic properties of simulated landﬁll components. ..................... 58

Table 2. Summary of recursive Qs and standard deviation in Qs for realizations
using heterogeneous and anisotropic waste properties. ....................... 63

PAPER NO. 4: Leachate Recirculation Using Horizontal Permeable Blankets in
BIOREACTOR Landﬁlls

Table 1. Saturated and unsaturated hydraulic properties used in HYDRUS-ZD
simulations. ..................................................................................... 105

PAPER NO. 5: GEOTECHNICAL SENSING SYSTEM TO MONITOR INJECTED
LIQUIDS IN LANDF ILLS

Table 1. Hydraulic conductivity of loose and dense crushed glass. ................ 151

PAPER NO. 6: LEACHATE RECICRCULATION IN BIOREACTOR LANDFILLS
USING GEOCOMPOSITE DRAINAGE MATERIAL

Table 1. Physical Properties of the Geocomposite Drainage Material. ........... 180

LIST OF FIGURES

INTRODUCTION

Figure 1. Simulated wetted patterns of leachate recirculation using an LRS
consisting of: (a) 3 HTS spaced 20 m; (b) 5 HTS spaced 15 m; and (c) a
60-m-wide PB for a total leachate injection rate of 5.4 m3/hr/m. ........ 4

Figure 2. Simulated wetted areas for LRS consisting of: (a) 3 HTS spaced 20 m;
(b) 5 HTs spaced 15 m; and (c) a 60-m-wide PB for a total leachate
injection rate equal to 2. 7 m 3/hr/m. ..................................................... 6

PAPER NO. 1: LEACHATE RECIRCULATION USING HORIZONTAL
TRENCHES IN BIOREACTOR LANDFILLS

Figure 1. Conceptual model for HYDRUS-ZD simulation of leachate
recirculation in MSW landﬁll using horizontal trench system (a); and
detail of a horizontal trench (b). ........................................................ 15

Figure 2. Simulated leachate ﬂux at steady-state versus leachate injection
pressure head for a 0. 6—m-wide by 1—m-deep single HT for kW— — 107,
10'6 ,and 10'5 m/S. ............................................................................. 23

Figure 3. Simulated wetted areas at steady-state for 0. 6-m-wide by 1.0—m-deep
single HT for kW = 10 6m/s for (a) H,-= 5 m (degree of saturation >
45% to 2 90%), and (b) H,- ranging from 0 to 5 m (degree of saturation
Z 90%). ............................................................................................ 24

Figure 4. Simulated leachate ﬂux at steady state versus leachate injection
pressure head as a function of width and depth of HT for kW- — 10'6 m/s.
......................................................................................................... 27

Figure 5. Distribution of wetted area corresponding to degree of saturation Z
90% at steady-state for LRS consisting of seven HTs at 10-m-
horizontal spacing (a); and ﬁve HTS at 15 m spacing (b) for kW - 1045
m/s and H;- — 1. 5 m ............................................................................ 29

Figure 6. ......... Effect of horizontal spacing between HTS and leachate injection
pressure head on the hydraulic efﬁciency for LRS consisting of two
HTsinarowforH=1.,5 3, and5mand kW=1O rn/s. .................. 31

Figure 7. (a) Simulated total leachate ﬂux at steady-state versus leachate
injection pressure head for LRS consisting of four, ﬁve, and seven HTS
spaced horizontally at s- - 10, 7.5, and 5 m for kW— *- 10'6 m/s; and (b)
hydraulic efﬁciency. ......................................................................... 32

Figure 8. Hydraulic efﬁciency as a function of vertical spacing for LRS

consisting of two vertically positioned HTS for H,- = 1.5 and 5 m and
kW=10'6m/s ..................................................................................... 34

xi

Figure 9. Hydraulic efﬁciency as a function of number of HTS positioned
vertically at a ﬁxed spacing of 5 m and for kW = 10*6 m/s. ................. 37

Figure 10. Pattern of wetted area at degree of saturation Z 90% at steady-state
for geometric formations A, B, C and D for H1: 3 m, kW = 104’ m/s, s =
20m, andD= 15m. .......................................................................... 38

Figure 11. Hydraulic efﬁciency for formations A, B, and C presented in Figure
10. .................................................................................................... 39

PAPER NO. 2: EVALUATION OF HETEROGENEITY AND AN ISOTROPY OF
WASTE PROPERTIES ON LEACHATE RECIRCULATION IN
BIOREACTOR LANDFILLS

Figure 1. Conceptual model for HYDRUS—ZD Simulation of leachate
recirculation in MSW landﬁll using horizontal trench system (a); and
detail of a horizontal trench (b) ......................................................... 57

Figure 2. Conceptual model for generation of a random hydraulic conductivity
realization using correlation lengths Xx and AZ, where 3,, > A1, and

hydraulic conductivity variance (0 2) is lognonnally distributed ....... 61

Figure 3. Distribution of wetted area corresponding to degree of saturation Z
90% at steady-state using homogeneous and isotropic properties of
waste having kw = 104 cm/S and H, = 2 m. ....................................... 65

Figure 4. Simulated recursive average leachate ﬂux versus number of
realizations for geostatistical realizations corresponding to Sets 1 and
2. ...................................................................................................... 67

Figure 5. Distribution of wetted area corresponding to degree of saturation 2
90% for geostatistical realizations corresponding to Set 1 (1,, = 10 m,
A, = 0.5 m, and o 2 = 0.42) (a); and Set 2 (71,, = 2 m, x, = 0.5 m, and o
2 = 0.42) (b) for lognormal average kw = 104 cm/s and H = 2 m ...... 69

Figure 6. Schematic of leachate breakout distance (a); and simulated leachate
breakout distance for simulations using Sets 1 and 2 for kW = 104 cm/s
and Hi = 2 m (b) ............................................................... 71

PAPER NO. 3: NUMERICAL EVALUATION OF PERMEABLE BLANKETS FOR
LEACHATE RECIRCULATION IN MSW LANDF ILLS

Figure l. (a) Conceptual model for simulation of leachate recirculation in MSW
landﬁll using a conventional horizontal trench and permeable blanket
alternative using HUDRUS-ZD; (b) detail for permeable blanket, (c)
detail for horizontal trench ................................................................ 80

xii

Figure 2. Mositure distribution for leachate recirculation system containing (a) 3
horizontal trenches; and (b) a permeable blanket. ............................. 81

Figure 3. Simulated steady state recirculated leachate ﬂux versus leachate
injection pressure head for single horizontal trench for kW = 10'6, 5 x
10‘, and 10'5 m/s. ............................................................................. 85

Figure 4. Simulated steady-state recirculated leachate ﬂux versus leachate
injection pressure head for single horizontal trench and for 30-m-wide
and 60-m-wide permeable blankets for kW = 10’6 m/s and km& k3 =
10'2 111/s. ........................................................................................... 87

Figure 5. Simulated steady-state recirculated leachate ﬂux versus leachate
injection pressure head for 60-m-wide permeable blankets having k3 =
103, 5x103, and 10'2 m/s. ................................................................. 89

Figure 6. Simulated steady-state recirculated leachate ﬂux versus leachate
injection pressure head for single horizontal trench, seven horizontal
trenches spaced at 10m c/c, and a 60-m-wide permeable blanket having
kg and km = 10'2 m/s. ....................................................................... 91

PAPER NO. 4: Leachate Recirculation Using Horizontal Permeable Blankets in
BIOREACTOR Landﬁlls '

Figure 1. (a) Conceptual model used for numerical simulation of leachate
recirculation in MSW landﬁll using a horizontal permeable blanket;
and (b) detail of a permeable blanket. ............................................. 103

Figure 2. Schematic of wetting front of injected leachate in permeable blanket.
............................................................................................................... 107

Figure 3. Effect of kW on simulated wetted width and pressure head of injected
leachate for a 60-m-wide PB for Q = 1.1 m3/hr/m (a); and Q = 3.6
m3/hr/m (b) ..................................................................................... 111

Figure 4. Simulated wetted width of waste and simulated pressure head in the
PB as a function of the soil waster characteristic curve used in
simulating the waste ......................................................... 113

Figure 5. Effectoka on simulated wetted width and pressure head of injected
leachate for a 60—m-wide PB for: (a) Q = 1.1 m3/hr/m; and (b) Q = 3.6
m3/hr/m. ......................................................................................... 114

Figure 6. Effect of PB depth on Simulated wetted width and pressure head of
injected leachate for a 60-m-wide PB. ............................................ 116

Figure 7. Simulated wetted width of waste as a function of leachate dosing

frequency for a 150-m-wide PB (a); and corresponding simulated
pressure head in the PB (b). ............................................................ 120

xiii

 

Figure 8. Contours of Simulated maximum wetted width of waste at steady-state
as a function of kg and kW for continuous leachate injection rates: (a) Q
= 1.1 m3/hr/m; and (b) Q = 3.6 m3/hr/m. ......................................... 121

Figure 9. Simulated wetted width of waste as a function of the initial degrees of
saturation of the waste and permeable blanket for a 150-m-wide PB.
....................................................................................................... 123

Figure 10. Simulated wetted width and pressure head of injected leachate as a
function of the degree of saturation of waste (SW) for a 60-m-wide PB.
....................................................................................................... 125

Figure 11. Simulated wetted width and pressure head of injected leachate versus
leachate injection period as a function of the initial degree of saturation
of blanket ($3) for a 60-m-wide PB ................................................. 126

Figure 12. Correlation between the wetted width of waste measured in the ﬁeld
and the simulated wetted width of waste for a 60-m-wide PB for kW
ranging from 10‘7 m/S to 10'5 m/s, s3 = 30%, Q = 1.1 m3/hr/m, SW =
45% (a); and SW = 65% (b). ............................................................ 130

Figure 13. Measured and simulated pressure head of injected leachate in the
glass permeable blanket for three leachate injection events

corresponding to Q = 1.1 m3/hr/m ................................................... 131
PAPER NO. 5: GEOTECHNICAL SENSING SYSTEM TO MONITOR INJECTED
LIQUIDS IN LANDFILLS
Figure 1. Schematic of a permeable blanket (a) perspective view; and (b) cross
section AA’. ................................................................................. 141 1
Figure 2. Geotechnical sensors used in this study. ....................................... 1466

Figure 3. Lab evaluation of electrical conductivity of liquid on the
measurements by TDR and impedance moisture content sensors. 154

Figure. 4. Field response of TDR and impedance moisture content sensors to a
leachate injection event. ................................................................. 157

Figure 5. Field response of pressure transducers and thermistors to a leachate
injection event. ............................................................................... 159

Figure 6. Field response of pressure transducers and impedance moisture
content sensors to a leachate injection event. .................................. 161

Figure 7. Effect of diurnal air temperature on the ﬁeld measurements of
thermistors and thermocouples ....................................................... 164

PAPER NO. 6: LEACHATE RECICRCULATION IN BIOREACTOR LANDFILLS

USING GEOCOMPOSITE DRAINAGE MATERIAL

xiv

Figure 1. Schematic of wetted area at steady-state simulated by HYDRUS-2D
for leachate recirculation system consisting of: (a) four horizontal
trenches; (b) a horizontal permeable blanket; and (c) a sloped
permeable blanket. ......................................................................... 175

Figure 2. Schematic of an instrumented GDL leachate recirculation blanket: (a)
plan view; and (b) cross section AA’ through the blanket in west — east
direction. ..................................................................................... 17878

Figure 3. Veriﬁcation of response of impedance moisture content sensor using
TDR moisture content sensor at location B 9 m. .......................... 18484

Figure 4. Migration of leachate in the GDL blanket evaluated by change in
temperature and increase in pressure head measured by embedded
sensor at location E 4.5 m. .............................................................. 186

Figure 5. Response of impedance moisture content sensors located in the (a)
eastern; and (b) western portions of the blanket to a leachate injection
event at leachate injection rate equal to O. 9 m 3/hr/m. ...................... 188

Figure 6. Effect of leachate injection rate on the rate of travel of leachate in the
blanket for leachate injection events corresponding to injection rates
equal to O. 9 and 2. 6 m /hr/m. .......................................................... 190

Figure 7. Correlation between simulated and measured rate of travel of injected
leachate for leachate injection events corresponding to injection rates
equal to 0. 9 m 3/hr/m. ...................................................................... 193

Figure 8. Measured and simulated pressure head of injected leachate in the GDL
blanket for two leachate injection events corresponding to Q= 0. 9
m 3/hr/m ........................................................................ 194

Figure 9. Leachate injection pressure and leachate injection rate measured at the
site. ................................................................................................ 197

Figure 10. Leachate pressure heads and temperatures within the blanket and
waste temperature outside the blanket measured at the site. ............ 198

XV

KEY TO ABBREVIATIONS

ID = internal diameter
D1 = deionized water
F G = leachate ﬂow gauge

GDL = gecomposite drainage layer

GDLS = more than a single geocomposite drainage layer
GN = geonet

GT = geotextile

HDPE = high density polyethylene

HT = horizontal trench

HTS = horizontal trenches

KCI = potassium chloride

LCS = leachate collection system
LR = leachate recirculation

LRS = leachate recirculation system

MSW = Municipal solid waste

PB = permeable blanket

PBs = permeable blankets

SWCC = soil water characteristic curve

SWCCS = soil water characteristic curves

TDR = time domain reﬂectometry

TDRs = more than a single time domain reﬂectometry

TSS = total suspended solids

xvi

KEY TO SYMBOLS

Co = coefﬁcient of curvature
Cu = coefﬁcient of uniformity
D = vertical spacing between horizontal trenches in a leachate recirculation

system or the vertical spacing between a horizontal trench and leachate
collection system

D50 = average particle size diameter
E 4.5 m = sensor location at 4.5 m east of the injection pipe
Gs = speciﬁc gravity
H,- = leachate injection pressure head
M = electrical impedance moisture sensor
N = total number of horizontal trenches in a leachate recirculation system
or number of realizations used to estimate recursive leachate ﬂux, Q,
NE 9 m = sensor location at 9 m northeast of the injection pipe
NW 4.5 m = sensor location at 4.5 m northwest of the injection pipe
P = piezometer
Q, = recursive average of Q5
Q3; = simulated leachate ﬂux at steady-state condition corresponding to the
ith realization
Q5 = Simulated leachate recirculated ﬂux at steady-state condition for a
single lHT '
Qm) = Simulated recirculated leachate ﬂux at steady-state condition for the
lower HT in an LRS consisting of 2 HTS positioned vertically
Qsm = simulated total recirculated leachate ﬂux at steady-state condition for
an LRS consisting of more than one HT
st) = simulated recirculated leachate flux at steady-state condition for the
upper HT in an LRS consisting of 2 HTs positioned vertically
R = impedance measured by moisture content sensor
S = sink term in Richards equation
Se = effective degree of saturation
SE 9 m = sensor location at 9m southeast of the injection pipe
SW 4.5 m = sensor location at 4.5 m southwest of the injection pipe
T = thennocouple
W 9 m = sensor location at 9 m west of the injection pipe
WW = wetted width of underlying waste
W3 = saturated wetted width of the permeable blanket
d = spacing between leachate collection system pipes
h,- = leachate injection pressure head
hp = leachate pressure head
i = iteration number or realization count
k3 = saturated hydraulic conductivity of the permeable blanket material
kHT = saturated hydraulic conductivity of HT backﬁll
kLcs = saturated hydraulic conductivity of the drainage material in leachate
collection system
kW = saturated hydraulic conductivity of MSW
kWH = saturated hydraulic conductivity of MSW in horizontal direction

xvii

 

Pa
02

= saturated hydraulic conductivity of MSW in vertical direction

= saturated hydraulic conductivity

= dimensionless ﬁtting parameter for van Genuchten (1980) ﬁtting
function for soil-water characteristic curve

= dimensionless ﬁtting parameter for van Genuchten (1980) ﬁtting
function for soil-water characteristic curve or porosity

= lateral spacing between horizontal trenches in a leachate recirculation
system

= time

= thickness of leachate collection system

= vertical spatial dimension through the landﬁll model domain

= ﬁtting parameter for van Genuchten (1980) ﬁtting function for soil-
water characteristic curve

= slope angle of leachate collection system with respect to horizontal

= unsaturated hydraulic conductivity as a function of matric suction head

= volumetric water content

= residual volumetric water content

= saturated volumetric water content

= hydraulic efﬁciency of leachate recirculation system

= matric suction head

= correlation length of waste hydraulic conductivity in the horizontal
direction

= correlation length of waste hydraulic conductivity in the vertical
direction

= dry density

= variance in kW for lognormal distribution

xviii

 

INTRODUCTION

BACKGROUND

In the United States, Resource Conservation and Recovery Act (RCRA) Subtitle D
regulations require containment, collection, and treatment of leachate before it is released
into the environment. Leachate recirculation (LR) is a leachate management alternative
that is commonly practiced in municipal solid waste (MSW) landﬁlls. LR operation
consists of injecting leachate into MSW and collecting it using leachate collection system
(LCS) located above the lining system at the bottom of the landﬁll. LR offers signiﬁcant
economical and environmental beneﬁts for MSW landﬁlls. These beneﬁts include: (1)
added ﬂexibility in leachate storage; (2) reduction in the leachate disposal and treatment
costs; (3) reduction in the discharge of leachate constituents to the environment due to
reduction in the volume of leachate treated and discharged by waste water treatment
plants; (4) accelerated decomposition and settlement of waste resulting in an airspace
gain and an increase in the rate of gas production; (5) faster improvement in the leachate
quality; and (6) potential reduction in the post-closure care period and associated
maintenance costs.

Currently, there are no speciﬁc design guidelines available for designing leachate
recirculation systems for MSW landﬁlls. Due to the absence of such design guidelines,
these risks and drawbacks cannot be appropriately evaluated: (1) potential decrease in the
factor of safety for slope stability of landﬁlls; (2) potential increase in the leachate head
on the liner if the leachate collection system is not designed to efﬁciently drain injected

leachate; (3) potential ﬂooding of the gas collection system; and (4) leachate seeps from

 

 

the sides of the landﬁll if an appropriate buffer distance is not provided. Hence, before
implementation, landﬁll operators should weigh the advantages and risks associated with

leachate recirculation on a site-speciﬁc basis.

CONVENTIONAL LEACHATE RECIRCULATION METHODS

Conventional leachate recirculation techniques can be divided into surface and subsurface
application. Surface application consists of: (1) direct application of leachate or spray
irrigation of leachate on the landﬁll surface; or (2) surface ponding of leachate. Surface
application is climate dependant, labor intensive, and requires surface availability.
Advantages of surface application include potentially uniform inﬁltration of leachate into
the waste and reduction in leachate volume due to a greater evaporation. Disadvantages
of surface application include odor problems, poor aesthetics, and potential runoff of
applied leachate into storm water management system.

Most commonly used conventional subsurface leachate recirculation systems
(LRSS) consist of horizontal trenches (HTS) and vertical wells. HTS are more commonly
used in relatively modern landﬁlls, while vertical wells are relatively common in retroﬁt
landﬁlls where it is not cost-effective or possible to install HTS. Unlike surface
application, subsurface application is not labor intensive and do not cause odors and
direct leachate exposure unless leachate seeps out from the Sides of the landﬁll.
Disadvantages of using conventional subsurface LRSs include: (1) higher capital cost for
the installation compared to surface application; and (2) need for an excavation during
construction resulting in odor problems.

Due to the lack of Speciﬁc design guidelines, design of the conventional systems is

done on an ad hoc basis with no clear understanding of the effect of design parameters on

 

the long-term performance of the LRS. Lack of speciﬁc design guidelines for subsurface
LR has resulted in non-uniform wetting of the waste causing: (1) leachate seeps resulting
in discharge concerns; (2) differential settlement resulting in increased cap maintenance
costs; (3) under-generation of landﬁll gas affecting gas-to-energy revenues; and (4) lack
of understanding to design and implement a monitoring system. Hence, in this
dissertation, one of the tasks consisted of performing a numerical parametric study to

provide speciﬁc design guidelines for subsurface LRSS

NEW LEACHATE RECIRCULATION METHOD
This dissertation evaluates a new subsurface LRS called “permeable blanket” as an
alternative LRS for MSW landﬁlls. The permeable blanket (PB) design concept consists
of placing a relatively thin and high hydraulic conductivity material on a relatively ﬂat or
inclined waste surface in a landﬁll. A perforated pipe is embedded in the PB in the
direction parallel to the shorter or longer plan view dimension of the PB where leachate is
injected under a positive pressure. The relatively high permeability of PB results in
preferential travel of injected leachate within the blanket before the leachate inﬁltrates
through the underlying waste.

The key advantages of PBS over conventional systems are: (1) excavation of waste
is not needed during the construction of the blanket resulting in less odor problems; (2) a
PB acts as an engineered heterogeneity and reduces the effect of spatial variation of waste
properties when wetting the waste resulting in a relatively uniform distribution of leachate
in the landﬁll; and (3) a PB can substitute multiple horizontal trenches (Figure 1) or

vertical wells requiring very little pipe work. PBS made up of granular materials can also

 

“ Dry Zones”

 

 

 
  
   
   

HTs @

O
20 m spacing \ z 90 /'

SaturatIon

 

 

 

 

 

 

 

L“ Low Saturatloh:
Pockets”

 

HTs @

“ Dry Zones”
1 5 m spaclng ﬂ

  
   

 

 

 

 

 

 

 

-- mo ,-
L_L_“ Low Saturation _l

Pockets”

 

BO-m-wlde
PB e“;

 

 

Z 90%
Saturatlon

  
   

 

 

 

 

 

 

o g
\ 5 “Low Saturation _/

Pockets”

Figure 1. Simulated wetted patterns of leachate recirculation using an LRS consisting of:
(a) 3 HTS spaced 20 m; (b) 5 HTS spaced 15 m; and (c) a 60-m-wide PB for a total
leachate injection rate of 5.4 m3/hr/m.

provide an ideal platform to embed sensors for monitoring the pressure, temperature, and
migration of injected leachate.

Figure 1 presents numerically simulated wetting patterns resulting from LR using
HTS (Figures 1a and 1b) versus that for a PB (Figure 1c). Heterogeneity and anisotropy in
the hydraulic conductivity of waste was simulated for the waste matrix in this evaluation.
The distribution of kW used for the waste matrix was maintained the same for all scenarios
presented in Figure 1 in order to allow comparison under equivalent conditions.

Total leachate injection rate equal to 5.7 m3/hr/m (cubic meter per hour per meter
length of the injection pipe) was simulated for each of the LRS presented in Figure 1 .The
wetted areas shown in Figure 1 correspond to a degree of saturation 2 90% at steady-state
condition. Figures la and 1b Show that an LRS made of HTS can result in the formation
of “dry zones”. Dry zones are the zones where injected leachate cannot reach and hence
cannot wet the waste. Reducing the lateral Spacing and increasing the density of HTS can
reduce the formation of dry zones. However, it is relatively expensive. Figures 1a and 1b
also present the presence of low saturation pockets. These pockets result from high kW
pockets that do not retain moisture. Figure 1c shows that an LRS made of a PB resulted in
a relatively more uniform distribution of injected leachate with less dry zones. However,
low saturation pockets similar to those observed for HTS were observed.

Leachate injection rate used for a PB should be chosen such that injected leachate
can ﬁll up the entire or required width of the PB without causing excessive pressure
buildup in the PB. If injected leachate cannot ﬁll up the design width of a PB, it is not
economical to use this PB due to costs associated with the excess material used for

constructing the PB. Figure 2 presents the wetting patterns resulting ﬁ'om LR using HTs

 

“ Dry Zones” _1

 

I-ITs @
20 m spacing

Leachate Collection

System _\

 

i l

.52.)

 
 
 
 

O

 

z 90%
Saturation

 

 

 

 

HTs @
15 m spacing

z 90%
Saturation

 

 

./—— “ Dry Zones”

 
  

 

Pockets”

  

 

\——L—“ Low Saturation —/

 

60-m-wide
PB We

 

 

Figure 2. Simulated wetted areas for LRS consisting of: (a) 3 HTS spaced 20 m; (b) 5
HTS Spaced 15 m; and (c) a 60-m-wide PB for a total leachate injection rate equal to 2.7

m3/hr/m.

   

 

z 90%
Saturation

 

 

\ 5 “ Low Saturation k

Pockets”

 

 

 

and a PB. The wetted areas shown in Figure 2 correspond to a degree of saturation 2
90% at steady-state condition. The leachate injection rate simulated in Figure 2 was equal
to 2.7 m3/hr/m, which is half of the injection rate used for the wetted areas presented in
Figure l. The distribution of kW used for the set of simulations presented in Figure 2 is
same as that for simulations presented in Figure 1. Figure 2c when compared to Figure 1c
shows that the wetted area for the PB was greatly reduced when the leachate injection
rate was reduced to half the value.

There are other concerns that need to be considered when designing an LRS using
PBS. Biological activities that occur in landﬁlls may partially or fully clog the PB and
decrease its hydraulic conductivity and hence may jeopardize the hydraulic efﬁciency of
the LRS. When a PB settles or sags, to compensate for the elevation head loss, the pump
needs to have the adequate head. If PBS are subjected to excessive non-uniform vertical
loads, differential settlement may jeopardize the physical integrity and the hydraulic
continuity of the blanket and cause disruption in the ﬂow. In addition, liquid pressure
heads generated in the blanket during leachate injection periods need to be closely
monitored to keep the factor of safety for slope stability of the landﬁll in an acceptable

range.

OBJECTIVES AND METHODOLOGY

The key objectives of this dissertation are to: (1) prepare design guidelines by performing
parametric studies for leachate recirculation or liquid addition using HTS and PBS; and
(2) evaluate the use of a PB to recirculate leachate for MSW landﬁlls. To achieve the ﬁrst

objective, numerical modeling of HTS and PBS was carried out. To achieve the second

 

objective, numerical modeling followed by ﬁeld-scale testing of three instrumented PBS
was carried out at an active MSW landﬁll located in Jackson, Michigan.

The saturated/unsaturated ﬂow model HYDRUS-ZD was used to perform the
parametric study. The key design parameters Simulated in the study were: leachate
injection rate or pressure, leachate dosing frequency, geometry of LRS, and hydraulic
properties of the waste and LRS.

For the ﬁeld-scale testing, three 60-m-long and 10-m-wide PBS made up of crushed
recycled glass, shredded tires, and a geocomposite drainage material were constructed.
Total about 50 sensors were embedded in these blankets to measure moisture content,
pore-water pressure, temperature, and vertical stress. Leachate was injected at ﬂow rates

ranging from 0.9 m3/hr to 3.6 m3/hr per unit meter length of the injection pipe.

DISSERTATION ORGANIZATION

This dissertation has been organized into six sections. Each section is written in the form
of a technical paper. The ﬁrst paper presents a parametric study, using numerical
modeling, for the design of HTS used for leachate recirculation. The second paper
addresses the effect of heterogeneity and anisotropy of waste hydraulic properties on the
performance of LRSS consisting of HTS. The third paper compares the use of HTS to PBS
for recirculating leachate in MSW landﬁlls using a steady-State modeling approach. The
fourth paper presents a parametric study for designing LRSS consisting of PBS including
ﬁeld data from the instrumented PB constructed using crushed glass. The ﬁﬁh paper
presents the design of an automated sensing system used in the ﬁeld to monitor pressure,

temperature, and travel of injected leachate in the PBS. The sixth paper presents the

 

results from ﬁeld-scale testing of the instrumented PB constructed using a geocomposite

drainage material.

PAPER NO. 1: LEACHATE RECIRCULATION USING
HORIZONTAL TRENCHES IN BIOREACTOR LANDF ILLS

ABSTRACT

Leachate recirculation in municipal solid waste (MSW) landﬁlls operated as
bioreactors offers signiﬁcant economical and environmental beneﬁts. Subsurface
leachate recirculation in MSW landﬁlls is commonly achieved by using horizontal
trenches or vertical wells. Currently, there are no design guidelines available for
leachate recirculation using subsurface leachate recirculation system (LRS). The key
objective of this study is to prepare design guidelines for LRS consisting of horizontal
trenches. This paper presents a numerical study of LRS consisting of horizontal
trenches. The design parameters evaluated in this study include: (1) leachate injection
pressure head; (2) hydraulic conductivity of trench backﬁll and MSW; (3) dimensions
of trench; and (4) spacing and geometric formation of trenches. The ﬁnite-element
saturated/unsaturated ﬂow model HYDRUS-ZD was used for the numerical study.
The hydraulic performance of the LRS was evaluated primarily using the simulated
recirculated leachate ﬂux and distribution of ﬂow under steady-state ﬂow condition.
The key ﬁndings of this numerical study are: (1) logarithm of leachate ﬂux and
leachate injection pressure head have a curvilinear relationship and leachate ﬂux is
directly proportional to the hydraulic conductivity of MSW; (2) if hydraulic
conductivity of trench backﬁll is equal to or greater than that of MSW, any further
increase in the hydraulic conductivity of the trench backﬁll has negligible impact on
leachate ﬂux; (3) for a given cross sectional area, horizontal trenches having width
greater than depth can recirculate greater leachate ﬂux and can wet more area of the

waste; and (4) reduction in the horizontal spacing between trenches and vertically

10

 

staggering the trenches reduces “dry zones” between trenches where otherwise

recirculated leachate may not reach.

INTRODUCTION
Leachate recirculation (LR) is a leachate management alternative that is commonly
used for municipal solid waste (MSW) landﬁlls. In the United States, Subtitle D
regulations (Resource Conservation and Recovery Act) require containment,
collection, and treatment of leachate before it is released into the environment. LR
operation consists of injecting leachate into MSW and collecting it using leachate
collection system (LCS) located above the lining system at the bottom of the landﬁll.
LR offers many environmental and economical beneﬁts (Doran 1999; Mehta et a1.
2002) to MSW landﬁlls including: (1) reduction in leachate treatment and disposal
costs; (2) greater ﬂexibility in leachate management and treatment; (3) relatively
faster biodegradation of waste resulting in increased gas production and quicker waste
stabilization and settlement; (4) reduction in the risk associated with contamination
from Spills during off-site transportation, treatment, and disposal of leachate; and (5)
potential reduction in the post-closure care period of the landﬁll. Trucking and
leachate treatment costs in the US. range from approximately $5 to $25/m3
(Leachator 2004). Depending upon the site-speciﬁc leachate quantities generated, LR
can save hundreds of thousands of dollars for a typical medium size landﬁll over its
design life.

Leachate recirculation also has a few disadvantages. These disadvantages
include: (1) reduction in the shear strength of MSW potentially reducing the factor of
safety against slope stability of the landﬁll; (2) potential leachate breakouts from the

sides of the landﬁll; and (3) increase in the leachate head build up on the liner

11

 

 

potentially increasing the risk for ground water contamination. Hence, designers and
landﬁlls owners are expected to weigh the advantages and disadvantages on a site-
Speciﬁc basis before a leachate recirculation system (LRS) is implemented.

Leachate recirculation can be performed using multiple techniques. These
techniques are divided into surface and subsurface application. Surface application
consists of: (1) direct application of leachate or spray inigation of leachate on the
landﬁll surface; or (2) surface ponding of leachate. Odor problems, poor aesthetics,
and potential runoff of applied leachate into storm water management system are key
the drawbacks of these surface application techniques.

Conventional subsurface application techniques are (Khire and Haydar 2003;
Qian et a1. 2002): (1) vertical wells; and (2) horizontal trenches (HTS). Unlike surface
application of leachate, subsurface LRS do not cause odors and direct leachate
exposure unless leachate seeps out from the sides of the landﬁll. HTS are more
commonly used in modern lined landﬁlls. Vertical wells are relatively common in
retroﬁt landﬁlls where it is not cost effective or not possible to install HTS. Currently,
there are no speciﬁc design guidelines available for designing subsurface LRS. The
key objective of this study is to prepare design guidelines for LRS consisting of HTS.

Various cross-sectional dimensions of HTS have been reported in the literature
including 0.6-m-wide by l-m-deep (Maier and Vasuki 1996), 1.0-m-wide by 1.0-m-
deep (Miller and Emge 1997), and 0.9 to 1.2-m-wide by 1.2 to 1.8-m-deep (Reinhart
and Carson 1993). HTS are backﬁlled with relatively high conductivity drainage
material (e.g. gravel, coarse sand, crushed glass, shredded tires) with a 50 to 150 mm
diameter perforated pipe installed at the center of the trench. Leachate is recirculated
by injecting leachate into the perforated pipe at positive pressures ranging from 0 to 5

m (SWANA 2002). Townsend and Miller (1998) have presented a ﬁeld study on

12

leachate recirculation using horizontal benches backﬁlled with shredded tires. Typical
horizontal spacing of HTS varies from 3 to 30 m (SWANA 2002). In this paper, in
order to meet the key objective, we have evaluated these parameters: (1) leachate
injection pressure head; (2) hydraulic conductivity of HT backﬁll; (3) dimensions of
HT; and (4) horizontal spacing and the geometric formation of HTS. We have
evaluated the effect of these parameters on the leachate ﬂux that can be recirculated at
steady-state (Q5) using the saturated/unsaturated ﬂow model HYDRUS-ZD developed

by Simunek et al. (1999).

NUMERICAL MODELING

HYDRUS-ZD Computer Model

HYDRUS-ZD is a computer model that can Simulate water, heat, and solute transport
in unsaturated, partially saturated, or fully saturated porous media (Simunek et al.
1999). The program numerically solves the Richards’ Equation for
saturated/unsaturated water ﬂow. A 2-D form of Richards’ equation can be expressed

as follows:

ﬂ€__£ 61-3 91”. M-
at ’ ax[k(wax] az[k(‘”)az + 62 S (I)

where 6 = volumetric water content; i// = matric suction head; k = hydraulic
conductivity of the porous material which is strongly dependant on the soil suction or
water content; 2 = vertical dimension; S = volume of water removed per unit time per
unit volume of soil by plant water uptake (sink term); and t = time.

HYDRUS-2D was selected for this study due to its diverse capabilities.
HYDRUS-ZD can be used to simulate ﬂow regions delineated by irregular
boundaries. The boundaries can be selected as constant or time-variable prescribed

head, ﬂux, or controlled by atmospheric conditions. The ﬂow region can be simulated

l3

with an arbitrary degree of local anisotropy and heterogeneity. The model also has a
built-in database for hydraulic properties of soils and can incorporate hysteresis in the
soil-water characteristic curves. The model can present the output in a graphical
format as plots or contours of water content, pressure head, and velocity. This model
has been used for saturated/unsaturated liquid and solute transport through porous
media in several studies (Scanlon et al. 2002; Henry et al. 2002; Pang et al. 2000;

Rassam et al. 2002).

Conceptual Model and Assumptions

A schematic of the conceptual model used to simulate the LRS is presented in Figure
1. We Simulated leachate as pure water in this study. Henceforth, any reference to
leachate ﬂow corresponds to water ﬂow. Nevertheless, the results of this study can be
applied to any liquids as long as the liquid has physical and hydraulic properties that
are relatively close to water. The effects of gas ﬂow, temperature, and biochemical
reactions occurring within a landﬁll were ignored. The maximum leachate injection
head simulated was 5 m to cover the maximum injection head that has been used in
the ﬁeld (SWANA 2002). Most existing landﬁlls where leachate is recirculated use
less than 5 m injection head to reduce leachate breakouts and to keep the factor of
safety against slope stability failure of the landﬁll in an acceptable range (SWANA
2002)

Korﬁatis at al. (1984) have demonstrated success when the authors applied
Richards’s equation for saturated and unsaturated ﬂow to a lab-scale MSW sample
having 0.5 m diameter and 1.5 m height. A wide variety (size and composition) of
organic and inorganic MSW deposited in landﬁlls results in MSW that exhibits

heterogeneity and anisotropy in its hydraulic properties. Hence, the soil-water

l4

Zero Flux Boundaries

 

\
~ MSW -
Lcs Pipe HT [See Detail in (0)] Zero Flux
(Seepage Face <— 3 -> Boundary
(a) Boundary) E
( Leachate Perforated LR Pipe
LCS L (Constant Head
Boundary)

 

d 60 m Liner (Zero Flux
Notes: Boundary)
HT = Horizontal Trench.
LCS = Leachate Collection System. (b) Depth

LR = Leachate Recirculation.
MSW = Municipal Solid Waste.
NOT TO SCALE.

  

9'9wa

HT Backﬁll

Figure 1. Conceptual model for HYDRUS-ZD Simulation of leachate recirculation in
MSW landﬁll using horizontal trench system (a); and detail of a horizontal trench (b).

characteristic curves for waste can vary signiﬁcantly. McCreanor (1998) simulated
the effect of daily cover on the spread of recirculated leachate and found that daily
cover having lower hydraulic conductivity than waste can enhance lateral Spreading
but restrict vertical spreading of leachate. The ﬂow pattern of leachate is also affected
by channeling (ZeiSS and Ugguccioni 1995). In this numerical study, we assumed
MSW as a homogeneous and isotropic porous medium. Effect of channeling was not
considered in this study. Even though this assumption may not be completely in line
with ﬁeld conditions, the results from this numerical study can be useﬁrl in comparing
designs or to investigate alternatives during an iterative design phase of an LRS
(Straub and Lynch 1982). The effect of heterogeneity and anisotropy of MSW on the
hydraulic performance of LRS consisting of HTS is presented in detail by Haydar and
Khire (2005). Haydar and Khire (2005) found that the introduction of heterogeneity
and anisotropy resulted in a greater leachate ﬂux compared to that when waste was
assumed homogenous and isotropic for an equivalent average hydraulic conductivity

of the waste.

Boundary Conditions and Mass Balance

All external boundaries were simulated as zero-ﬂux boundaries (Figure 1). Leachate
flow as a result of percolation from the cap or waste above the model domain was
assumed zero. We believe that this assumption is reasonable because our key
objective in this study was to evaluate the subsurface hydraulics of recirculated
leachate. The perforated pipe used for leachate injection was simulated as a constant
head boundary (Figure 1b). The diameter of the leachate injection pipe was assumed
equal to 0.1 m based on what is commonly used in the ﬁeld. The leachate injection

pressure heads ranged from zero (gravity drainage) to 5 m. The leachate injection

16

pressure head assigned as constant head was exclusive of head loss in pipes, joints,
manifolds, and pumps used in a typical LRS.

We also conducted a limited number of simulations by assigning the perforated
leachate injection pipe as a constant ﬂux boundary. These Simulations yielded the
same constant head at the injection pipe for the corresponding ﬂux at steady-state.
Thus, assigning the leachate injection pipe as a constant head or constant ﬂux
boundary yields the same result for steady-state ﬂow condition.

Leachate collection pipes embedded in the LCS were simulated as seepage face
boundaries. The diameter of such LCS pipes was assumed equal to 0.15 m. The
minimum Size of the ﬁnite-elements used for discretization of the problem domain,
the time step, and the error tolerances for pressure head and water content were
selected such that cumulative water balance error did not exceed 0.1%. In order to
achieve such a low mass balance error, the problem domain was divided into
triangular ﬁnite elements having maximum dimensions ranging from 1 mm to 25 mm.
We used an error tolerance of 0.1% for the volumetric water content and 10'4 m for
the matric suction. We used a minimum time step of 8 X 1045 second and a maximum
time step of 8 hours. HYDRUS-ZD gradually increases the time step automatically if
mass balance and error tolerance criteria are met. Typically it took about 1 to 2 days

for completing a simulation on a Pentium 2.5 MHz processor.

Hydraulic Properties

Three material types were selected for Simulating the key components of the LRS.
These three components included MSW, HT backﬁll, and the LCS. The three
materials for these components were simulated as homogeneous and isotropic porous
materials. The materials and the saturated hydraulic conductivities of these materials

are presented in Table 1.

l7

Table 1. Saturated and unsaturated hydraulic parameters input to HYDRUS-2D to
simulate leachate recirculation system consisting of horizontal trenches.

 

 

 

 

 

 

 

a ks
Landﬁll Unit Material 0, 6, (1 /m) n (mls)
MSW Silt loam 0.067 0.45 2 1.41 105, 10‘, & 10'7
Pea gravel 0.01 0.3 57.44 2.44 10'2
HT Backﬁll
Sand 0.045 0.43 14.5 2.68 10“
LCS Pea gravel 0.01 0.3 57.44 2.44 10'2

 

 

 

 

 

 

Notes: MSW = municipal solid waste; HT = horizontal trench; LCS = leachate

collection system.

18

The hydraulic conductivities of MSW, kW, were selected based on the typical
values published by Hughes et al. (1971), Fungaroli and Steiner (1979), Korﬁatis et
al. (1984), Oweis et al. (1990), and Bleiker et al. (1993). The hydraulic conductivities
of the HT backﬁll, km, were selected based on the values for pea gravel, shredded
tires, or coarse sand which are commonly used as the HT backﬁll material (Doran
1999). The hydraulic conductivity of the LCS drainage material, kLcs, was selected
based on the values for pea gravel which is commonly used for constructing LCS of
MSW landﬁlls. The unsaturated hydraulic properties for the materials listed above
consisted of the van Genuchten (van Genuchten 1980) ﬁtting parameters for the soil-
water characteristic curves and unsaturated hydraulic conductivity function (Table 1).
The soil-water characteristic curves for the materials simulated are represented bythe

van-Genuchten model (van Genuchten 1980) as follows.
as _ 6r

(1 new)“

 

9=0r+ (2)

where y! = matric suction head; 6 = water content; 63 = saturated water content; 6, =
residual water content; and a, n, and m (m =1 -n” ) are ﬁtting parameters. HYDRUS-
2D model uses the van Genuchten-Mualem (Mualem 1976) function to predict the
unsaturated hydraulic conductivities using the van Genuchten ﬁtting parameters and
the saturated hydraulic conductivities. The unsaturated hydraulic conductivities are
estimated by using the van Genuchten-Mualem model (Mualem 1976) presented in
Equation 3.

1 m 2
__ 0.5 —
k(W)-ksSe 1—[1“Sem] (3)

19

where Se is effective degree of saturation; k, is saturated hydraulic conductivity; k(1//)
is unsaturated hydraulic conductivity function; in is the matric suction; and m is the
ﬁtting parameter for the soil-water characteristic curve.

The unsaturated hydraulic properties for the materials were selected from the
database built into HYDRUS-ZD for soils having saturated hydraulic conductivities
closest to the assumed saturated hydraulic conductivities. Except for the saturated
hydraulic conductivity, rest of the parameters presented in Table 1 do not have any
inﬂuence on the leachate ﬂux predicted by the model at steady-state ﬂow condition.
However, transient leachate ﬂux before the steady-state is reached is a function of the
initial conditions and all saturated and unsaturated hydraulic parameters presented in
Table 1. The time to reach steady-state is primarily a function of the initial moisture
content, the injection pressure head, and the unsaturated hydraulic properties of the
waste. The unsaturated hydraulic properties also inﬂuence the water content proﬁle in
the capillary zone of the LRS where degree of saturation is less than 100%. In this
paper we have presented the leachate ﬂux after the steady-state ﬂow condition was

reached. Transient analysis of leachate recirculation is beyond the scope of this study.

Dimensions

The dimensions of the Simulated rectangular HT ranged from 0.6- to 2-m-wide by 1-
to 2-m-deep. These dimensions were selected based on the most common designs
used at existing landﬁlls (Miller and Emge 1997; Doran 1999). The thickness of the
LCS layer, ths, was assumed equal to 0.3 m. The US. EPA Subtitle D regulations
require that the leachate head on the liner not exceed 0.3 m. The hydraulic
conductivity of the LCS drainage material, krcs, was assumed equal to 10'2 m/s. The
spacing between the adjacent leachate collection pipes, d, was assumed equal to 60 m

(Figure 1a). The slope (tan/3) of the LCS was assumed equal to 3.5%. These LCS

20

design parameters (ths, kLCS, d, and tanﬂ) resulted in less than 0.3 m leachate head on
the liner for all simulations conducted in this study. The leachate head on the liner
was continuously monitored using observation points distributed along the liner. The
vertical distance, D, between the row of HTS and the top of the LCS was assumed
equal to 15 m. When this distance was varied in the model, we found that leachate
ﬂux at steady-state was not affected for distances greater than or equal to 3 m.
Distance from the top of the horizontal trench to the upper zero ﬂux boundary was
assumed equal to 8 m to contain all leachate ﬂow and to prevent artesian conditions

under the simulated leachate injection heads of up to 5 m.

MODELING RESULTS

In this study, over 100 simulations using HYDRUS-ZD were conducted to evaluate
the effect of the design parameters on the steady-state leachate ﬂux, Q,. We assumed
the leachate ﬂux reached steady-state when the injected leachate ﬂux equated the total
leachate ﬂux seeping from the LCS pipes located within the LCS (Figure 1a). Note
that Q, in m3/d/m represents the leachate ﬂux that can be recirculated at steady-state
in cubic meters per day per linear meter length of the HT perpendicular to the plane of
the paper.

Leachate recirculation in the ﬁeld is often carried out in on/off dosing cycles.
Hence, steady-state ﬂow condition can be rarely achieved in the ﬁeld. In the ﬁeld,
when the LRS is turned off, gravity drainage of leachate creates storage space in the
voids of MSW. Until steady-state is reached, leachate ﬂux greater than Q, can be

recirculated in the landﬁll for a given injection head.

21

Q, represents the lower limit of leachate ﬂux that can be recirculated in a
landﬁll for the given set of design parameters. Hence, a LRS designed using the Q,

values presented in this manuscript would be conservative.

Effect of Leachate Injection Pressure Head

The simulated leachate injection pressure head, Hi, ranged from 0 to 5 m. The HT
Simulated to evaluate the effect of injection pressure head was 0.6-m-wide by l-m-
deep. Figure 2 shows a curvilinear relationship between H; and log Q3. The shape and
approximate dimensions of the wetted area correspond to degree of saturation ranging
from 45% to 90% at steady—state for H,- = 5 m and kW = 10'6 m/s are presented in
Figure 3a. Figure 3a shows that the wetted area corresponding to a lower degree of
saturation is larger. Figure 3b presents the Shape and approximate dimensions of the
wetted area at steady—sate for degree of saturation 2 90% and for H; ranging from O to
5m. Figure 3b shows that as injection head increases the wetted area corresponding to
degree of saturation 2 90% at steady-state expands and hence Q, increases.

To illustrate the wetted area at steady-state, we have selected degree of
saturation Z 90%. However, we are not endorsing the idea of saturating waste to 90%
during leachate recirculation. Target degree of saturation to achieve optimum
bioreactor performance is beyond the scope of this study. We believe that the degree
of saturation to achieve optimum bioreactor perforamcne will vary based on
composition, density, organic fraction, waste temperature, and meteorological factors.
We recommend design engineers and landﬁll operators to select a target degree of
saturation based on site-Speciﬁc factors. Our modeling results Show that the shape and
size of the wetted area corresponding to degree of saturation 2 90% are independent

of kW. This ﬁnding is similar to saturated flow through homogeneous and isotropic

22

 

100 ~-.r.-I...J-..I.-.I..-I.”I.-.'....I...I--.1...1...I....L..T...I....I...7...I...I.-.I...I....I....I.....

L ............................................................................................... 4
C1223:II:iS:IIIZIIIICIIIIIIIIII33:1::I:::::::::I:II:.‘::I23:2:333:3:3333333331:13:3:
................... .................. .
j...........e..en...1....--..-o—o.-.o...:. .................. 5 t .................. .1
,. .................. ' ................... ' ................... = ' .: .................. ..
_ ____________ Bachus et al. (2002) ................ kw 1° "VS\ ______________

 

 

 

    

; ; 'Doran'(1999) & -----------------
~ ------------------ g ------------------- g ----------- R.W. Beck (2002)-mg ----------------- 3

 

Steady-State Leachate Flux. Q (m’ldlm)

 

 

 

 

(,0,
0 1 2 3 4 5

Leachate Injection Pressure Head, H, (m)

Figure 2. Simulated leachate ﬂux at steady-state versus leachate injection pressure
head for a 0.6-m-wide by l-m-deep single HT for kW = 107, 10 , and 10-5 111/5.

23

 

 

 

 

 

 

 

HTﬁ
(a) 2 457
2 GOV 6 m
2 70V
2 80V 13 m
2 90?
Li‘: 19 m ‘1‘
22 m ,
‘— 25 m —’l
HT——\
“3.2 m

 

 

 

 

 

 

Hi: 1.5m

(b) H1=5m
H,-=3m \\ 4)
13m

 

 

 

Figure 3. Simulated wetted areas at steady-state for 0.6-m-wide by 1.0-m-deep single
HT for kW = 106 m/S for (a) H,- = 5 m (degree of saturation 2 45% to 2 90%); and (b)

H,- ranging from 0 to 5 m (degree of saturation 2 90%).

24

porous materials analyzed using a ﬂow net. The shape of the ﬂow net is independent
of the hydraulic conductivity of the homogeneous and isotropic soil

For a given Hi, Q, is directly proportional to kW. Thus, if kW is increased or
decreased by an order of magnitude, assuming all other parameters remain constant,
Q5 would also increase or decrease by an order of magnitude, respectively.
McCreanor and Reinhart (1999) have reported leachate recirculation rates of 2 to 8
m3/d per linear meter of HT. These values correspond to Q, that we simulated for kW
ranging from 5 x 10'6 to 1045 m/s and H,- ranging from 0 to 5 m. In order to validate the
modeling results presented in Figure 2, we compared the simulated Q, to the values
for HTS measured in the ﬁeld by Doran (1999) and to that Simulated by Bachus et al.
(2002). We have discussed the validation of our modeling results in detail in a section

entitled “Model Validation” in this manuscript.

Effect of Hydraulic Conductivity of Trench Backﬁll

HTS are typically backﬁlled with granular material such as gravel, crushed recycled
glass, shredded tires, or coarse sand. The backﬁll is used to enhance the distribution
of leachate, to protect the perforated leachate injection pipe by providing a bedding
layer for a relatively uniform support, and to prevent clogging of the pipe from ﬁnes
in the surrounding MSW. For a 0.6-m-wide by 1.0-m-deep HT, we varied km from
104 to 10'2 m/s, kW was varied from 1045 to 10'5 m/S and H.- was varied from zero to 5
m. The simulations indicated no signiﬁcant impact on the magnitude of Q, and on the
shape of the wetted area when km was equal to or greater than kW. For example, when
kW was 10'6 m/s and km was equal to 10'6, 10's, 10'4 m/s or more, it resulted in about
the same Q3. However, when km was reduced below 1045 m/s, Q, decreased. Because
the cross sectional area of an HT represents a relatively small area fraction and MSW

represents a much larger area fraction of the model domain (Figure 1), until km drops

25

below kW, the equivalent hydraulic conductivity of the model domain does not alter

signiﬁcantly to impact Qs. Henceforth, for all simulations, we used km = 10'2 m/s.

Effect of Horizontal Trench Geometry and Size

We evaluated the effect of HT size (width and depth) by maintaining kW constant at
104’ m/S and varying the width and depth from 0.6 to 2 m. Four combinations of HT
sizes were simulated. The results are presented in Figure 4. Figure 4 Shows that Q,
increases with increase in the cross-sectional area of the HT. The increased cross
sectional area of HT results in a wider wetted area and hence a greater leachate
wetting volume. Greater leachate wetting volume results in greater Q3. However, for a
given area of an HT, width greater than depth results in greater Q3. This ﬁnding is also
conﬁrmed by Khire and Haydar (2003) where the authors have demonstrated that
relatively thin but laterally extensive blankets can recirculate relatively large leachate
ﬂux. The difference in Q, for the four combinations of trench sizes presented in
Figure 4 is relatively small. This is because the cross sectional area of HT represents a
relatively small area fraction of the model domain (Figure 1). Henceforth, for all

simulations we used 0.6-m—wide by 1.0-m-deep HTS.

Effect of Horizontal Trench Spacing

The horizontal Spacing, s, between adjacent HTS is one of the key parameters in LRS
design. It affects the volume, duration, and frequency of leachate dosing cycle for the
LRS. If the HT spacing is reduced, it translates into additional number of trenches for
a given coverage area, which increases the capital cost. For a given injection pressure
head, the spacing required to achieve degree of saturation Z 90% to virtually eliminate
dry zones between adjacent trenches can be estimated by comparing the wetted area

corresponding to degree of saturation 2 90% at steady-state. A dry zone is that region

26

N
—4
a
-—<
——4

 

   

g r
i .-

"E i : 'J
vats ~W'dth (In) ”Depth (m) _: ________________ 3|
0 = 2.0 x 0.6 g i

 

 

 

Steady-State Leachate Flux,

 

O i 2 3 4 5
Leachate Injection Pressure Head, H, (m)

Figure 4. Simulated leachate ﬂux at steady state versus leachate injection pressure
head as a function of width and depth of HT for kW = 107’ 111/8.

27

of MSW where recirculated leachate may not reach. Figure 5a presents the wetted
area corresponding to degree of saturation _>_ 90% for H,- =1.5 m at steady-state for s =
10 m.

Figure 5b presents the wetted area corresponding to greater than 90% saturation
for H,- =1.5 m at steady-state for s = 15 m. At IO-m-spacing, the wetted areas
corresponding to degree of saturation _>_ 90% do overlap. Hence, at spacing equal to or
less than 10 m, dry zones would not result or would be relatively small. However, at
15 m or greater Spacing, the wetted areas corresponding to degree of saturation 2 90%
do not overlap. Hence, relatively large area of dry zones would result for spacing
equal to or greater than 15 m for H,- = 1.5 m. If H,- is increased, the area of dry zones
would shrink.

For LRS consisting of two or more HTS, we have used the indicator parameter
hydraulic efﬁciency (27),) to quantify the hydraulic efﬁciency of the LRS and the
relative area of dry zones. This parameter can be used to optimize the LRS design.
We have computed 77;. by dividing the total flux of LRS achieved at steady-state
[Qsm] by the product of number of HTS in the LRS, N, and the leachate ﬂux achieved

at steady-state for LRS consisting of a single HT, Q,, as presented in Equation 4.

_ Qs(T)

— NxQS (4)

77h

Hydraulic efﬁciency, as deﬁned by equation 4, can never exceed 100%. If 77;, is equal
to l or 100%, it means that the HT spacing is such that wetted areas corresponding to
degree of saturation Z 90% do not overlap for a given Hi. Thus, there is a good
likelihood that dry zones would result. If 77;, is less than 100%, wetted areas
corresponding to degree of saturation 2 90% will overlap for a given Hi. Hence, there

would be less likelihood for dry zones to result. As 77;, further decreases, the wetted

28

HTs spaced 2 90°/ saturation
at 10 m l o

 

(a)

 

 

 

 

 

 

 

 

60 m
/—Dry zones

V

HTS spaced
at15

‘b’ Q Will - lateral

Figure 5. Distribution of wetted area corresponding to degree of saturation > 90% at
steady-state for LRS consisting of seven HTS at 10- m-horizontal spacing (a); and ﬁve
HTS at 15 m Spacing (b) for kW=10 m/S and H=1.5 n1.

 

 

 

 

 

 

 

29

areas would further overlap and this will result in a more uniform wetting. However,
it will decrease the hydraulic efﬁciency of the LRS (Eq. 4) and will increase the
capital cost associated with greater number of HTS required for the LRS. If the key
objective of the LRS is to minimize dry zones, the spacing shall be selected such that
m. is slightly less than 100%.

For a given kW and for km > kW, 771,15 a function of s and Hi. Figure 6 shows that
reducing 3 beyond a certain spacing for a given H,- reduces 77).. If H,- iS increased, s can
be increased without increasing the area of dry zones. Hence, for a given s, a greater
H.- can result in a lower value of 77;, because the overlap of the wetted areas
corresponding to degree of saturation z 90% increases (Figure 6).

We Simulated s = 5, 7.5 and 10 m over a 30 m width resulting in 7, 5, and 4
total HTS respectively. We maintained kW constant at 10'6 m/s. Figure 7a shows a
relatively small increase in total Q, [or Qs(7]] as s is reduced and N is increased. Figure
7b shows a decrease in 77;. as s is decreased. AS 3 is reduced, wetted area
corresponding to degree of saturation 2 90% for adjacent HTS begins to overlap. This
reduces the dry zones and reduces 77;. for the LRS. Data presented in Figure 7
combined with the shape and Size of the wetted area presented in Figure 3b for an
individual HT can be used to optimize the horizontal spatial distribution of HTS to
achieve a relatively uniform wetting of the MSW.

Al-Yousﬁ et al. (1992) presented an analytical approach to estimate the
maximum Spacing for horizontal perforated leachate injection pipes of an LRS
consisting of HTS under the ﬁnal cap of a sanitary landﬁll. The authors proposed
Equation 5 for the maximum distance between consecutive perforated injection pipes:

0.5
s = 211,811] (5)

WV

30

 

h

P
to

Hydraulic Efficiency, r) H
.0
oo

 

 

 

i

 

.0] l i l
0 5 10 15 20 25

Horizontal Spacing, s (m)

Figure 6. Effect of horizontal spacing between HTS and leachate injection pressure
head on the hydraulic efﬁciency for LRS consisting of two HTS in a row for H, = 1.5,
3, and 5 m and kW= 10*6 WS.

31

 

(ma/dim)
01

0
cm
.b

 

 

 

 

Total Steady-State Leachate Flux,

 

 

 

 

 

0
0 1 2 3 4 5
Leachate Injection Pressure Head, H, (m)
1 T. l T l
c: .............. ....... Q; ....... E ............................... T
: ' 4HTS
ats=10m ;
3 0.9 ‘ .. .............. 1
>2
2 i thT375m p
2 08 C ----------------------------------- -- _:
0 .
E . j
I“ " 1
g 0.7 — ......................................................................... P
:i “ . .
E E ti 53 ; E
1’ : qua 0" ‘ 5 mi 1
5‘ 0,5 _-(Al-Yousﬁ etal.1992).;. - ....... g ....... atsfs m ,,,,,,, _
:ibi i 3 .
0.5 1 l 1 1
0 1 2 3 4 5

Leachate Injection Pressure Head, H, (m)

Figure 7. (a) Simulated total leachate ﬂux at steady-state versus leachate injection
pressure head for LRS consisting of four, ﬁve, and seven HTS spaced horizontally at s
= 10, 7.5, and 5 m for kW= 10'6 m/s; and (b) hydraulic efﬁciency.

32

where s is the distance between horizontal injection pipes; H,- is the leachate injection
head; and km and kwy are the horizontal and vertical hydraulic conductivities of
waste, respectively.

For homogeneous and isotropic MSW, hydraulic conductivity is equal in all
directions (ka = kwy). Hence, Equation 5 converts to: s = 2 x H,. Equation 5 is plotted
in Figure 7(b) for s = 5, 7.5, and 10 m. The plot of Equation 5 in Figure 7b indicates
that m, is about 60%, 75%, and 84% when HTS are spaced at 5, 7.5, and 10 m and
operated at H,- = 2.5, 3.8, and 5 m, respectively. These values of 17;. are far below
100%. Hence, if HTS are Spaced based on Equation 5, wetted areas of the adjacent

HTS would overlap signiﬁcantly when H,- < 5 m. Hence, design based on Eq. 5 would

be extremely conservative.

Effect of Vertical Spacing

We evaluated the effect of vertical spacing, D, between HTS by simulating two HTS
separated vertically by a distance, D = 1, 5, 10, 15 and 30 m for kW = 1045 m/s and H;
= 1.5 and 5 m. Figure 8 presents 77;, as a function of D for H,- = 1.5 and 5 m. Figure 8
shows that as D is reduced, 77/, decreases. For a given value of D, 77;. increases as H,-
increases (Figure 8). We also observed that simulated leachate ﬂux at steady-state for
the upper HT [Q3(U)] is greater than that for the lower HT [Qsm]. Table 2 presents
these ratios: Q3“), / [gm], + Qs(L) ] and QWL) / [stj + QWL) ] for H,- = 5 m and for D
ranging from 1 to 30 m. Note that Q34), + Q3”, )= Qsm- AS D is reduced, the overlap of
the wetting front of the HTS increases. This leads to decrease in Qs(7)- Table 2 and
Figure 8 show that once the vertical Spacing is 15 m or more, 77;, of the LRS reaches a
constant value of about 85% for H,- = 5 m. At lower values of H1, D corresponding to

77;. = 85% is greater than 15 m (Figure 8).

33

 

.0 .O
\1 00

Hydraulic Efﬁciency, '71. H
.o
O)

 

 

 

 

0.5 ‘
0 5 10 15 20 25 30

Vertical Spacing, D (m)

Figure 8. Hydraulic efﬁciency as a function of vertical Spacing for LRS consisting of
two vertically positioned HTS for H.- = 1.5 and 5 m and kW = 1043 m/s.

34

Table 2. Effect of vertical spacing between HTS on the hydraulic efﬁciency of LRS.

 

 

 

 

 

 

 

 

Vertical Spacing, i 9% i gﬂ E
D (m) Qsir) Q50) 2 x Q3

1 0.65 0.42 0.53

5 0.83 0.57 0.70

10 0.91 0.61 0.76

15 0.97 0.62 0.80

30 0.99 0.64 l 0.82

 

 

 

35

Figure 9 presents 77;, as a function of H,- for LRS consisting of two and three
HTS spaced vertically at D = 5 m. Figure 9 shows that, for a given value of Hi, 77;,
decreases as total number of HTS increases. This occurs because there is a greater

overlap of wetted areas of adjacent HTS.

Effect of Geometric Formation of HTS

It is common practice to install HTS in individual cells of a landﬁll at various
elevations depending upon the permitted height of the cell and ﬁlling schedule. We
simulated four possible geometric formations of HTS that are or could be used in LRS
to evaluate the shape and size of wetted areas and the effect on 771.. These formations
are presented in Figure 10. The vertical distance, D, between two consecutive rows
(R1, R2, or R3) of HTS, was ﬁxed at 15 m. H,- was varied from zero to 5 m and kW
was ﬁxed at 1045 m/s. Horizontal spacing, s, was set at 20 m for all rows of HTS
presented in Figure 10. Figure 10 Shows the wetted area at steady-state for degree of
saturation 2 90% for H, = 3 m. Figure 11 shows 77;. for formations A, B, and C.

Figure 10 indicates that vertically staggered HTS (Formations A and C) resulted
in the least area of dry zones. Figures 5 and 6 Show that if horizontal spacing between
HTS is less than or equal to 10 m, the wetted areas of individual HTS corresponding to
Z 90% degree of saturation overlap and hence minimize dry zones. Formation D in
Figure 10 shows that at 20 m horizontal spacing, there would be signiﬁcant area of
dry zones for H,- = 3 m. However, F onnations A and C in Figure 10 show that if HTS
can be installed at various elevations, it allows us to double the horizontal spacing
between HTS for a given row to about 20 m as long as the HTS are sta ggered. Figure
11 Shows that 77;. for Formations A and C is greater than that for Formation B. Figure

11 shows that if Formations A or C are to be implemented, a horizontal Spacing of 20

36

 

 

0.65 h

Hydraulic Efﬁciency, an H
O
O)

0.55

 

 

 

0.5

 

0 1 2 3 4 5 6
Leachate Injection Pressure Head, H (m)

Figure 9. Hydraulic efﬁciency as a function of number of HTS positioned vertically at
a ﬁxed Spacing of 5 m and for kW = 1045 m/s.

37

 

 

    

 

   

 

.81 Waited area for
- Z 9070 _-
‘ D j saturation Dry zones
R: v . HT
41
D
R, 1 .

 

 

 

 

 

 

 

 

 

 

Formation Formation B

 

 

Dry zones

Mn

 

 

 

 

 

 

 

 

Formation Formation D

Notes:

R1=Row1,R2=Row2.andR3=Row3

D = Vertical distance between two successive
rows of HTs

s = center to center spacing between HTS

Figure 10. Pattern of wetted area at degree of saturation Z 90% at steady-state for
geometric formations A, B, C and D for H,- = 3 m, kW = 10" m/s, 3 = 20 m, and D = 15
m.

38

 

1 l l T F

 

      
  

 

 

 

ne—ﬁ :
d FormationC -
0.95 h ....z
3 l 2
1:: 0.9 f ”FormationA- ‘ T
5‘ i i
5 : 4
'5 0.85 ~ . z
E : e 1
“‘ - j
o

g 0.8, 1
in FormationB I
R

0.7 i i 1 1 1
O 1 2 3 4 5

Leachate injection Pressure Head,H, (m)

Figure 11. Hydraulic efﬁciency for formations A, B, and C presented in Figure 10.

39

m at an injection head of 2 to 3 m would be sufﬁcient to signiﬁcantly reduce dry
zones. Thus, vertical staggering of HTS allows to increase the horizontal spacing

between HTS for a given row.

MODEL VALIDATION

The numerical study presented in this paper is primarily based on numerical modeling
conducted using HYDRUS-ZD. MSW is a highly heterogeneous and anisotropic
material consisting of pore ﬂuid (leachate) having complex geochemical properties.
Heterogeneity and anisotropy of MSW may not ever be fully characterized or the cost
of characterization using conventional ﬁeld methods would exceed the beneﬁts of
such characterization in design or operation of a MSW landﬁll. Hence, it may not be
possible to measure the representative values of water content and hydraulic
properties of MSW (Oweis et al. 1990). Due to these reasons, majority of the
modeling results presented in this paper have not or cannot be accurately validated in
the ﬁeld for MSW landﬁlls. There have been two modeling studies related to the
design of LRS consisting of HTS conducted by McCreanor (1998) and Bachus et al.
(2002). Doran (1999) has presented ﬁndings from a ﬁeld study of an LRS consisting
of HTS. In this section we have presented the results from these studies as a validation

of our modeling work.

McCreanor (1998)

McCreanor (1998) used the United States Geological Survey’s (USGS) Saturated-
Unsaturated Flow and Transport (SUTRA-ZD) model to simulate leachate
hydrodynamics using LRS consisting of a Single HT. The author simulated a 2-m-
wide by l-m-deep HT. The trench backﬁll had a porosity of 0.3 and hydraulic

conductivity equal to 10'3 m/s. The initial degree of saturation of MSW was assumed

40

equal to 40%. MSW was Simulated as a homogeneous and isotropic medium using kw
= 105, 10'6 and 10'7 m/s. McCreanor (1998) studied the effect of kWon wetted leachate
area by modeling the lateral and vertical Spread of the wetted area corresponding to
degree of saturation 2 45% a week after the leachate injection began.

McCreanor (1998) modeled the lateral and vertical movement of leachate by
simulating both, continuous and intermittent (8hr on/ 16hr off) application of leachate.
The modeling results showed that the lateral and vertical spread of wetted area is
almost similar for continuous or intermittent applications. The lateral and vertical
spread of the wetted area was measured along horizontal and vertical lines passing
through the centre of the perforated leachate injection pipe installed inside the HT.
We conducted similar modeling using HYDRUS-ZD and compared our modeling
results to the results presented by McCreanor (1998) for continuous application of
leachate at 2, 4, and 8 m3/d/m for one week. The wetted area dimensions
corresponding to degree of saturation of 45% or more simulated by HYDRUS-ZD and
SUTRA-ZD at the end of one week are presented in Table 3. The wetted area
dimensions simulated by both models are relatively close. We believe that the Slight
difference is because McCreanor (1998) used power function to ﬁt the soil-water
characteristic curves for the materials including MSW and we used van Genuchten

ﬁtting parameters in HYDRUS-ZD.

Bachus et al. (2002)

Bachus et al. (2002) used VS2DI (Hsieh et al. 2000) to simulate the liquid ﬂow in a
LRS consisting of HTS. VS2DI is a ﬁnite-difference model that simulates liquid ﬂow
and solute or energy transport in variably saturated porous media. The authors
simulated 0.9-m-wide by 0.9-m-deep HT and assumed homogenous and isotropic

material properties. The authors assumed kW= 10'5 m/s and H,- = 2.4 m and presented

41

Table 3. Comparison of simulation results from SUTRA-2D and HYDRUS-ZD.

 

 

 

 

 

Application and Rate Lateral Spread] SYII'Ztaitciadf
Flow Model of Leachate Injection 0f Wetted Area Wetted Area1
(m) (m)
SUTRA-ZD by Continuous
McCreanor (1998) Applicatiop at 2, 4 and 2'1’ 2'8 and 3'5 0'2’ 0'3 and 1'5
HYDRUS-ZD 8 m /d/m
(this study) (Time period ._. 1 week) 2.0, 3.4 and 5.3 0.4, 0.5 and 1.3

 

 

 

Note: 1. Corresponds to degree of saturation 2 45%.

42

the lateral and vertical Spread of wetting front corresponding to various degrees of
saturation for intermittent (8 hrs on and 16 hours off) leachate recirculation.

Bachus et al. (2002) also simulated steady-state leachate ﬂux for kW= 10'5 m/S
at various values of Hi. We have plotted Q, values presented by Bachus et al. (2002)
in Figure 2. The VSZDI results are relatively close to what we simulated using

HYDRUS-ZD.

Doran (1999)

Doran (1999) recirculated leachate using an LRS consisting of HTS at the Crow Wing
County MSW landﬁll (CWCL). The size of the horizontal trenches installed ranged
between 0.65-m-wide by 0.65-m-deep to 1.4-m-wide by 1.4-m-deep. The trenches
were backﬁlled with shredded tires. A system of seven laterals was used to recirculate
leachate under gravity drainage (H, = 0). However, these HTS had positive elevation
head ranging from about 1 m to 5 m. One lateral was run at a time and the recirculated
leachate ﬂux was recorded using ﬂow meters. Laterals 3 and 7 had a diameter of 0.1
m, were 150-m-long and embedded in 0.65-m-wide by 0.65-m-deep HTS. The
leachate ﬂux estimated by HYDRUS-ZD was compared to that obtained from
monitoring laterals 3 and 7. This ﬁeld leachate ﬂux measured at CWCL (R.W. Beck
2002) is plotted against the elevation head (~ injection pressure) in Figure 2. Note
that we ignored the head loss due to friction in the Doran (1999) and R.W. Beck
(2002) data in Figure 2. However, as per our estimates using Moody diagram (Moody
1944), the head loss due to friction in the laterals is less than 0.5 m. Figure 2 shows
that the recirculated leachate ﬂux at CWCL mostly fell within the range estimated by
HYDRUS-ZD for kW = 106 m/s. Doran (1999) has not measured the ﬁeld kW ,
however, the interpreted kW is in the range reported in the literature for kW (Korﬁatis

et al. 1984; Fungaroli and Steiner 1979; Hughes et al. 1971).

43

SUMMARY AND PRACTICAL IMPLICATIONS

This paper presents a numerical study of key design variables used for the design of

an LRS consisting of HTS. The design of the leachate pipe network including the head

loss in the pipes, joints, manifolds, and pump has not been considered in this study.

The design parameters evaluated included: (1) H,-; (2) km and kW; (3) geometry and

size of HT; and (4) Spacing and formation of trenches. The saturated-unsaturated ﬂow

model HYDRUS-ZD was used to simulate the hydraulics of leachate recirculation in

MSW landﬁlls operated as bioreactors. The key ﬁndings of this study are as follows.

log Q, and H,- have a curvilinear relationship;

When km is equal to or greater than kW, any further increase in km has negligible
impact on Q,;

Q, is directly proportional to kW for a given Hi;

If the width of HT is increased, it results in greater wetted area for a given degree
of saturation. Depth of HT has a smaller impact on wetted area compared to the
width of HT;

In order to optimize number of HTS in LRS and reduce dry zones between
adjacent trenches, 77;, deﬁned in this paper can be used to establish maximum
Spacing between 2 or more HTS. For HTS Spaced horizontally, hydraulic
efﬁciency slightly less than 1 (100%) is an indicator of relatively uniform wetting
of MSW and hence relatively fewer dry zones.

When two or more rows of HTS are designed, HTS in these rows should be
vertically staggered to reduce dry zones.

Because the lateral and vertical spreading of leachate under steady-state

conditions is not a function of kW, the shape and dimensions of the wetted area

44

presented in Figure 3b can be used to design formations of HTS to optimize and to
achieve relatively uniform wetting of MSW.

0 Based on the data presented in Figures 3b, 5, and 10, HTS should be placed at
about 10 m horizontal spacing for formations consisting of HTS in a Single row
and operated at H - S 5 m. HTS Should be placed at about 20 m horizontal spacing
for formations consisting of HTS in multiple rows where HTS are staggered (e. g.,
Formations A or C in Figure 10) and operated at H,- S 5 m.

o The Simulated Q, values presented in Figure 2 can be used to estimate the total
number of HTS, the rate of leachate recirculation, duration, and frequency of
leachate dosing. The leachate ﬂux values presented in Figure 2 depend on kW
Because it is almost impossible to measure representative values of ﬁeld kW and
kW may vary signiﬁcantly in a landﬁll, the designer needs to assume a range of
possible values for kW and incorporate redundancy in the design.

Although not explored in this study, we recommend that LRS designers consider the

effect of these issues when designing LRS: (1) potential increase in the leachate head

on the liner; (2) potential decrease in the shear strength of MSW affecting the slope
stability of the landﬁll; (3) potential leachate breakouts from the side Slopes of the
landﬁll; and (4) effect on the efﬁciency of gas collection system. We also recommend
that designers consider the effect of hydraulic properties of daily cover, potential of
clogging of LRS and the effect of waste settlement on the long-term performance of

the LRS.

ACKNOWLEDGEMENT
The authors are grateful to Dr. Jirka Simunek for his input related to running the

computer model HYDRUS-ZD. The results and opinions presented in this manuscript

45

are those of the authors. The authors also wish to thank the three anonymous

reviewers for the constructive comments.

46

REFERENCES

Al-Yousﬁ, B.A., Pohland, F.G., and Vasuki, NC. (1992). “Design of landﬁll leachate
recirculation systems based on ﬂow characteristics.” Proceedings of the 4 7'”
Purdue University Industrial Waste Conference Proceedings, Purdue
University, West Lafayette, Indiana, 191-200.

Bachus R., Jaber J., and Harris J. (2002). “Guidance for bioreactor design.”
Proceedings of the Abstracts ﬁom 2'"! International Landﬁll Research
Symposium, October, Asheville, NC, 59-60.

Bleiker, D.E., McBean, E., and Farquhar, G. (1993). “Refuse sampling and
permeability testing at the Brock West and Keele Valley Landﬁlls.”
Proceedings of the Sixteenth International Madison Waste Conference, 548-567.

Doran, F. (1999). “Lay leachate lay.” Waste Age, April, 74-79.

Fungaroli, A. and Steiner, R. (1979). “Investigation of sanitary landﬁll behavior.”
Vol. 1, Final Report, US. EPA 600/2-79-053a.

Haydar, M.M. and Khire, M.V. (2005). “Evaluation of heterogeneity and anisotropy
of waste properties on leachate recirculation in bioreactor landﬁlls.” Journal of
Solid Waste Technology and Management, in press.

Henry, E.J., Smith, IE, and Wanick, AW. (2002). “Two-dimensional modeling of
ﬂow and transport in the vadose zone with surfactant-induced ﬂow.” Water
Resources Research, Vol. 38, No. 11, 331-3316.

Hughes, 6., Landon, R., and Farrolden, R. (1971). “Hydrogeology of Waste Disposal
Sites in Northeastern Illinois.” EPA Solid Waste Management Series, SW-12d.

Hsieh, P., Wingle, W., and Healy, R. (2000). “VS2DI-A graphical soﬂware package
for Simulating ﬂuid ﬂow and solute or energy transport in variably saturated
porous media.” US. Geological Survey Water-Resources Investigations Report
99-4130, 16 p.

Khire, M.V. and Haydar, M.M. (2003). “Numerical Evaluation of Granular Blankets
for Leachate Recirculation in MSW Landﬁlls.” Proceedings of the Ninth
Sardinia Solid Waste Conference, Cagliary, Italy, Oct.

Korﬁatis, G., Demetracopoulos, A., Bourodimos, E., and Nawy, E. (1984). “Moisture
transport in a Solid Waste Column.” Journal of Environmental Engineering,
Vol. 110, No. 4, 789-796.

Leachator (2004). www.solidwaste.com.

Maier, T. and Vasuki, N. (1996). “Expected beneﬁts of a full-scale bioreactor
landﬁll.” Proceedings of Wastecon 1996, Portland, USA, 179-185.

47

McCreanor, P. (1998). “Landﬁll leachate recirculation systems: mathematical
modeling and validation.” Ph.D. Dissertation, University of Central Florida,
Orlando, FL.

McCreanor, P. and Reinhart, D. (1999). "Hydrodynamic modeling of leachate
recirculating landﬁlls." Waste Management and Research, Vol. 17, 465-469.

Mehta, R., Barlaz, M., Yazdani, R., Augenstein, D., Bryars, M., and Sinderson, L.
(2002). “Refuse Decomposition in the Presence and Absence of Leachate
Recirculation.” Journal of Environmental Engineering, Vol. 128, No. 3, 228-
236.

Miller, D. and Emge, S. (1997). “Enhancing landﬁll leachate recirculation system
performance.” Practice Periodical of Hazardous, Toxic, and Radioactive Waste
Management, July, 1 13-1 19.

Moody, L.F (1944). "Friction factor for pipe ﬂow." ASME Transactions, Vol. 66, 671.

Mualem, Y. (1976). “A new model for predicting the hydraulic conductivity of
unsaturated porous media.” Water Resources Research, Vol. 12, 513-522.

Oweis, 1., Smith, D., Ellwood, R., and Greene, D. (1990). “Hydraulic characteristics
of municipal refuse.” Journal of Geotechnical Engineering, Vol. 116, No. 4,
539-553.

Pang, L., Close, M., Watt, J., and Vincent, K. (2000). “Simulation of picloram,
atrazine, and simazine leaching through two New Zealand soils and into
groundwater using HYDRUS-ZD.” Journal of Contaminant Hydrol. , Vol. 44,
No. 1, 19-46.

Qian, X., Koerner, R., and Gray, D. (2002). “Geotechnical Aspects of Landﬁll Design
and Construction.” Prentice Hall, New Jersey.

R.W. Beck (2002). “2002 annual report: facility SW-376.” Crow Wing County
MMSW Landﬁll.

Rassam, D., Cook, F. (2002). “Numerical simulations of water ﬂow and solute
transport applied to acid sulfate soils.” J. Irrig. Drain. Eng, Vol. 128, No. 2,
107-115.

Reinhart DR. and Carson D. (1993). “Experiences with full-scale application of
landﬁll bioreactor technology.” The 31” Annual Solid Waste Exposition of the
Solid Waste Association of North America. San Jose. California, Aug.

Scanlon, B., Christrnan, M., Reedy, R., Porro, I., Simunek, J., and Flerchinger, G.
(2002). “Intercode comparisons for Simulating water balance of surﬁcial

sediments in semiarid regions.” Water Resources Research, Vol. 38, No. 12,
5901-5915.

48

Simunek, J., Sejna, M., and M.Th. Van Genuchten (1999). “The HYDRUS-2D
software package for Simulating the 2—D movement of water, heat, and multiple
solutes in variably Saturated media Version 2.0,” US. Salinity Laboratory,
Agriculture Research Service, USDA, Riverside, California.

Straub, W. and Lynch, D. (1982). “Models of landﬁll leaching: moisture ﬂow and
inorganic strength.” Journal of Environmental Engineering, Vol. 128, 231-250.

SWANA (2002). “The solid waste manager’s guide to the bioreactor landﬁll.”
SWANA, Oct.

Townsend, T.G. and Miller, W.L. (1998). “Leachate Recycle Using Horizontal
Injection,” Advances in Environmental Research, Vol. 2, No. 2, 129-138.

van Genuchten, M.Th. (1980). “A closed-form equation for predicting the hydraulic
conductivity of unsaturated soils.” Soil Science Society of America Journal, Vol.
44, 892-898.

Zeiss, C. and Uguccioni, M. (1995). “Mechanisms and patterns of leachate ﬂow in

municipal solid waste landﬁlls.” Journal of Environmental Systems, Vol. 23,
No.3, 247-270.

49

PAPER NO. 2: EVALUATION OF HETEROGENEITY AND
ANISOTROPY OF WASTE PROPERTIES ON LEACHATE
RECIRCULATION IN BIOREACTOR LANDFILLS

ABSTRACT

Leachate recirculation in municipal solid waste (MSW) landﬁlls was modeled using
HYDRUS-ZD, a saturated/unsaturated ﬂow model. The recirculation system consisted
of three horizontal trenches spaced laterally at 20 m. The model was used to simulate
the effect of heterogeneity and anisotropy in the hydraulic conductivity of MSW on
the leachate ﬂux that can be recirculated at steady-state ﬂow condition. Monte Carlo
analysis (MCA) was used for two sets of correlation lengths and variance using a
lognormal hydraulic conductivity distribution for MSW. MCA estimated the mean
leachate ﬂux and its standard deviation. MCA quantiﬁed the variability and the
uncertainty in the magnitude of leachate ﬂux. The key ﬁnding is that increasing the
correlation length ratio of hydraulic conductivity of MSW (greater stratiﬁcation)
results in relatively lower leachate ﬂux that can be recirculated at steady-state.
Heterogeneity and anisotropy in hydraulic conductivity of MSW also results in
relatively non-unifonn wetting of MSW. Hence, greater than 15 m buffer distance
between the landﬁll side slopes and the leachate injection system is needed to

minimize the potential of leachate breakouts from the side Slopes of the landﬁll.

INTRODUCTION

Leachate recirculation is commonly used in MSW landﬁlls because it offers these
economical and environmental beneﬁts (Khire and Haydar 2005; Miller and Emge
1997; Mehta et al. 2002): (1) reduction in leachate treatment and disposal costs; (2)

greater ﬂexibility in leachate management and treatment; (3) relatively faster

50

biodegradation of waste resulting in increased gas production and quicker waste
stabilization and settlement; (4) reduction in the risk associated with contamination
from Spills during off-site transportation, treatment, and disposal of leachate; and (5)
potential reduction in the post-closure care period of the landﬁll. Trucking and
leachate treatment costs can add up signiﬁcantly. Depending upon the Site-speciﬁc
leachate quantities generated, leachate recirculation can save hundreds of thousands
of dollars for a typical medium Size landﬁll over its design life.

Leachate recirculation also has few disadvantages. These disadvantages include:
(1) reduction in the shear strength of MSW potentially reducing the factor of safety
against slope stability of the landﬁll; (2) potential leachate breakouts from the sides of
the landﬁll; and (3) increase in the leachate head build up on the liner potentially
increasing the risk for ground water contamination. Hence, designers and landﬁlls
owners are expected to weigh the advantages and disadvantages on a site-speciﬁc
basis before a leachate recirculation system (LRS) is implemented. Preferential ﬂow
paths in MSW increase the potential for leachate breakout. In this numerical study, we
have presented the non-unifonn wetting in heterogeneous and anisotropic MSW due
to leachate recirculation.

Conventional subsurface application techniques for leachate recirculation are
(Khire and Haydar 2003; Qian et al. 2002): (1) vertical wells; and (2) horizontal
trenches (HTS). Currently, there are no Speciﬁc design guidelines available for
considering the effect of heterogeneity and anisotropy of MSW properties when
designing a LRS using HTS. Haydar and Khire (2005) have presented guidelines for
designing a LRS consisting of HTS for homogeneous and isotropic MSW. The key
objective of this study is to evaluate the effect of heterogeneity and anisotropy of

MSW properties when designing a LRS consisting of HTS.

51

HYDRAULIC PROPERTIES OF MSW

Particle size of MSW constituents ranges over many orders of magnitude due to the
presence of materials having orders of magnitude different dimensions, such as food
waste, plastic bags, carpet, plastic sheeting, white appliances, etc. This results in
signiﬁcant heterogeneity and anisotropy in the hydraulic conductivity, kW, and other
physical properties of MSW. Such variation can exist not only among landﬁlls, but
also within a landﬁll cell. Furthermore, as waste undergoes settlement, hydraulic
properties of waste may change. Thus, spatial and temporal variations in kW are
abundant in a typical MSW landﬁll.

Leachate recirculation does not wet the waste matrix uniformly due to the
heterogeneity in waste and also due to the hydraulic limitation(s) of recirculation
system(s) used (Haydar and Khire 2005). A non-uniform distribution of moisture
content in MSW landﬁlls can result in non-uniform biodegradation of waste. This can
inﬂuence the operational performance of a typical landﬁll intended to be an efﬁcient
bioreactor. Many ﬁeld investigations have reported a spatial variation of moisture
content within the waste (McCreanor and Reinhart 1998; Zomberg et al. 1999; Oweis
et al. 1990). Zeiss and Uguccioni (1995) found that channeling results in the
downward movement of leachate through interconnected pores at rates faster than the
uniform wetting front under Darcian ﬂow. The presence of preferential ﬂow paths
leads to an increase in the bulk hydraulic conductivity of waste. Even though we have
insufﬁcient data on the ﬂow of leachate through channels as waste settles due to
degradation or due to increased vertical stress, we expect that hydraulic conductivity
of MSW will decrease under such conditions.

We believe that the deposition and compaction of waste in relatively thin lifts

leads to horizontal stratiﬁcation within a landﬁll. The horizontal hydraulic

52

conductivity of the waste in landﬁlls has been estimated to be an order in magnitude
higher than the vertical hydraulic conductivity (Powrie and Beaven 1999). The
application of daily and intermediate soil covers adds to the anisotropy and
heterogeneity within a landﬁll. Field-scale tracer studies in landﬁlls have shown that
signiﬁcant amount of tracer was transported laterally out from the landﬁll compared
to that in the vertical direction (Rosqvist and Destouni 2000). This is due to
preferential ﬂow of leachate through horizontal layers. Leachate is expected to ﬂow
through channels created by the heterogeneous nature of waste and through
preferential paths created by stratiﬁcation of waste and the presence of horizontally
oriented relatively impervious materials such as plastic foils (Fellner et al. 2003).
Vertical stress in a landﬁll increases with the depth of waste. Hence, as depth
increases, waste density increases and hydraulic conductivity decreases (Bleiker et al.
1995).

The aim of this study is to evaluate the effect of heterogeneity or spatial
variation and anisotropy in hydraulic conductivity of MSW on (1) the leachate ﬂux
that can be recirculated in a landﬁll at a steady-state ﬂow condition; and (2) leachate
ﬂow patterns.

Waste in landﬁlls is spread and compacted in lifts with repeated passages of
compactor to reduce the volume of voids. AS per the US. federal regulations, at the
end of the working day, the waste placed in the landﬁll cell is required to be covered
using a lSO-mm-thick soil or an equivalent material to reduce odors and control
vermin and pests. For the model we used to simulate the spatial variation of waste, the
correlation lengths in the lateral and vertical direction were used to discretize the
Spatial distribution of kW. The thickness of waste lift and the process of spreading and

compacting waste dictate the correlation lengths of kW. In addition, the lateral

53

dimensions, thickness, and hydraulic properties of daily cover placed on waste also
inﬂuence the correlation lengths and the Spatial variation in hydraulic properties.
However, there are no published values for the correlation lengths for kW.
Measurement of kW in the ﬁeld that is representative of the waste matrix is relatively
challenging and expensive. In addition, due to signiﬁcant heterogeneity and
anisotropy of waste, kW measured for one landﬁll cell may not be representative of
another cell in the same landﬁll

Fenton and Grifﬁths (1996) studied the effect of spatially random soil hydraulic
conductivity on the seepage ﬂow rate and surface drawdown of an earth dam. The
earth darn was simulated as a heterogeneous and isotropic porous medium. Monte
Carlo analysis (MCA) was conducted for a spatially random ﬁeld using a lognormal
distribution for the soil hydraulic conductivity with prescribed mean, variance, and
correlation length. Simulation results showed that the mean seepage ﬂow rate

increased as the correlation length increased.

HORIZONTAL TRENCH LEACHATE RECIRCULATION SYSTEM

Various cross-sectional dimensions of HTS have been reported in the literature
including 0.6 m wide by 1.0 m deep (Maier and Vasuki 1996), 1.0 m wide by 1.0 m
deep (Miller and Emge 1997), and 0.9 to 1.2 m wide by 1.2 to 1.8 m deep (Reinhart
and Carson 1993). Unlike surface application of leachate, HTS do not cause odors
and direct leachate exposure unless leachate seeps out from the sides of the landﬁll.
HTS are backﬁlled with relatively high hydraulic conductivity drainage material (6. g.
gravel, coarse sand, crushed glass, Shredded tires) with a 50 to 150 mm diameter
perforated pipe installed at the center of the trench. Leachate is recirculated by

injecting leachate in the perforated pipe at positive pressures ranging from 0 to 5 m

54

(SWANA 2002). Typical horizontal Spacing of HTS varies from 3 to 30 m (SWANA
2002). In this paper, in order to meet the key objective, we have evaluated the effect
of heterogeneity and anisotropy in the hydraulic conductivity of MSW on the leachate
ﬂux that can be recirculated under a steady-state ﬂow condition (Q,). We have
conducted over 50 numerical simulations using HY DRUS-ZD model which is capable
of generating heterogeneity and anisotropy for given values of vertical and horizontal
correlation lengths and standard deviation for a lognormal distribution of hydraulic

conductivities.

METHODOLOGY

Leachate recirculation in MSW landﬁlls has often been modeled using 2-D models
assuming that the hydraulic properties of waste are homogeneous and isotropic. This
assumption is overly simplistic. Due to signiﬁcant heterogeneity of waste, various
approaches have been considered to describe the effect of waste heterogeneity on the
leachate ﬂow. The effect of anisotropy and heterogeneity of kW on the lateral and
vertical moisture distribution was modeled by McCreanor and Reinhart (1998). They
found that compaction may promote anisotropy in kW, which extenuates horizontal
spreading and retards the downward movement of leachate. Bendz (1998) used a
kinetic wave model to describe the channeling of leachate ﬂow through waste.

In this study we have conceptualized a model domain to simulate the LRS
consisting of horizontal trenches in MSW landﬁlls. The ﬁnite-element model
HYDRUS-ZD was used to Simulate the hydraulics of leachate ﬂow. The waste was
simulated as a homogenous and isotropic as well as a heterogeneous and anisotropic

porous material using two sets of geostatistical realizations.

55

Conceptual Model

Figure 1 presents the conceptual model used to Simulate leachate recirculation using
horizontal trenches for a MSW landﬁll. The model domain represents a vertical cross-
section through a typical landﬁll cell. The LRS consists of three 0.6 m wide by 1.0 m
deep trenches, backﬁlled with gravel and installed at 20 m horizontal Spacing. The
trenches are located 15 m above the liner and 5 m below the landﬁll surface. A
perforated pipe having 100 mm diameter is installed at the center of each trench to
inject leachate for recirculation. The gravel backﬁll has a saturated hydraulic
conductivity, km = 1 crn/s. The leachate collection system (LCS) consists of 0.3 m
thick gravel layer with two 150-mm-diameter perforated pipes embedded at 60 m
horizontal spacing (Figure l). The detailed hydraulic properties of the Simulated
landﬁll units are presented in Table 1. All external boundaries were assumed as zero-
ﬂux boundaries. The pressurized pipe in the trench is modeled as a constant pressure

head boundary. The LCS pipes were modeled as seepage face boundaries.

Numerical Model

The ﬁnite-element ﬂow model HYDRUS-ZD was selected for this study due to its
capability to simulate saturated/unsaturated ﬂow. Other key reasons for selecting
HYDRUS-2D are: (1) it can be used to implement a scaling procedure to approximate
hydraulic variability in a given soil proﬁle; (2) it simulates such variability by means
of a set of lognormal scaling transformations, which relate the individual soil
hydraulic characteristics to those of a speciﬁed reference soil; (3) it is supported by an
interactive graphics-based interface for data-preprocessing, generation of a structured
mesh, and post-presentation of the results; and (4) it also includes a parameter

optimization algorithm for inverse estimation of a variety of soil hydraulic parameters

(Simunek et al. 1999).

56

Zero Flux Boundaries

, ‘.' . r-i. . ' . . . . . . ._ .. '.'.' . 7.,\ .,. .. .. ... ... ... ... 1.. ... .(. t" I." x". v _..
. ,-. .13.; .-..' .- ../.-..». .-, , v . . ..~. .‘ .e. ... ... ... ..\. ... .-- .\. .3. ... H. .4. .‘-,‘«_a,_ .. .-
,.. .. .7- .4. . r. ‘ z. ,, . .. .\ . ,-- .. _..\ .5. ‘ .‘K‘ . .. . .\~ .\, .. .. .n -. .‘\ J. ,.
p’:_r./..,.,. ..I..:-. tr,» \, - . . ..w. .-._~. _ .. -. .. .. .,.\-\\ ‘3": . - \.\\\‘\. .
.. . . ... .. '.' .. “... ... . ... 3.x .. g . ... .. .. ..‘1.K . \. . ,. n . . .‘\. , t \
. ., .. I . .‘ ' ,‘ -‘ . . .' .1 - .—‘. . ' ' ‘ ‘ f ' e n s e o n n u ,
,. . , 1.. . . - .. \ N \ \ V
. .’ 1' 1 . .. .. r \ O . . ..‘ \ . \ ,
.°.4.‘.,.. .(v.“,/.\r I Saar \ , \ . .i -\ . a H w . - n \ r\ J .
e - x. . . . .r. r . . o \i. l \ ‘.. \ ax . \ . . n s V.

_. n ) t l r .
l'.\' - . o .v a
v . I - u n

. n
n“ l . t \
. .
I” l
.
. u

(a)

 

Zero Flux
Boundary

Perforated LR Pipe
(Constant Head
Boundary)

 

 

Liner (Zero
otes: Flux Boundary)
. HT = Horizontal Trench.

    

LCS = Leachate Collection System. (b)

N

1

2. Depth
3. LR = Leachate Recirculation.

4

5

MSW = Municipal Solid Waste.
NOT TO SCALE.

HT Backﬁll

Figure 1. Conceptual model for HYDRUS-ZD simulation of leachate recirculation in
MSW landﬁll using horizontal trench system (a); and detail of a horizontal trench (b).

57

Table 1. Hydraulic properties of simulated landﬁll components.

 

 

 

Landfill Units Material 0r 6, a (l/m) n k (cm/s)
Waste Silt 16am 0.067 0.45 2 1.41 10“
HT Backﬁll &

Leachate Gravel 0.01 0.3 57.4 2.4 1

Collection System

 

 

 

 

 

 

 

 

 

58

In HYDRUS-2D, the ﬂow of leachate through waste is modeled as ﬂow through
any porous material by using the Richards’ equation (Richards 1931) presented

below,

ﬁn 6k(iI/)_
a: - V[k(W)VV/l+ 62 S (1)

where 0 is volumetric water content;u1 is matric suction head; k(w) is the unsaturated
hydraulic conductivity function for the porous material which is a function of soil
suction head; z is vertical dimension; S is sink term, and t is time. The soil-water
characteristic curves for the materials Simulated are represented by the van-Genuchten

model (van Genuchten 1980) as follows,

a, —6i,

(1+Iai/II" )m

where w = matric suction head; 0 = water content; 0, = saturated water content; 0r =

 

9=9r+ (2)

residual water content; and or, n, and m (m =1-n'l) are ﬁtting parameters. HYDRUS-
2D model uses the van Genuchten-Mualem (Mualem 1976) function to predict the
unsaturated hydraulic conductivities using the van Genuchten ﬁtting parameters and
the saturated hydraulic conductivities. The unsaturated hydraulic conductivities are
estimated by using van Genuchten-Mualem model (Mualem 1976) presented in

Equation 3,

I m
kw) = kW 530-5 1— [1 — 5,5,7] (3)

where Se is effective degree of saturation; kW is saturated hydraulic conductivity; k(\|l)
is unsaturated hydraulic conductivity function; and m is the ﬁtting parameter for the

soil-water characteristic curve.

59

Throughout this numerical study, we assumed the reference hydraulic
conductivity of waste, kW, equal to 10'4 cm/s. kW was selected based on the typical
values published by Hughes et al. (1971), Fungaroli and Steiner (1979), Korﬁatis et
al. (1984), Oweis et al. (1990), and Bleiker et al. (1993). Hydraulic conductivities of
HT backﬁll and LCS drainage material (km and hes) were assumed equal to 10'2
cm/S. km and [ties were selected based on the values for pea gravel, shredded tires, or
coarse sand which are commonly used materials for the HT backﬁll and LCS drainage
layer (Doran 1999). The unsaturated hydraulic properties for waste were selected
from HYDRUS-ZD database corresponding to a soil having saturated hydraulic

conductivity closest to the assumed kW of 10” cm/s.

Monte Carlo Analysis

The spatial variation in hydraulic conductivity was simulated by generating
stochastically lognormally distributed scaling factors. This technique assumes that the
hydraulic variability in a given area can be approximated by means of a set of linear
scaling transformations that relate the individual hydraulic conductivity to the
reference characteristics. This technique is similar to the one introduced by Miller and
Miller (1956) and is also explained in detail by Vogel et al. (1991).

Figure 2 presents a schematic of the conceptual model of a random hydraulic
conductivity realization simulated by HYDRUS-2D, where the correlation length in
the horizontal direction, W.,, is greater than the correlation length in the vertical
direction, 12. A lognormal distribution was used for the scaling factors generated for
the hydraulic conductivity. The Spatial correlation was accommodated using

correlation lengths in both horizontal and vertical direction (x and z directions in the

60

 

4 100m b

 

 

HT Injection pipe kw is correlated within the area
determined by 4,, and ,1,

 

 

2x
1: ﬂ; 1 l
r

W

22m

 

   

W
A V

v v

 

 

 

 

 

./—' Scaling factors of kw
WWW} (Function of 02)

0.001k 0.1kw 10kw 100kw

Figure 2. Conceptual model for generation of a random hydraulic conductivity
realization using correlation lengths A, and 2.2, where it, > 31, and hydraulic
conductivity variance (0 2) is lognormally distributed.

61

2-D waste domain). Depending upon the type, composition, and compaction methods,
we believe that the hydraulic conductivity of waste in a landﬁll cell varies. We
assumed 3., to range from 2 m to 10 m. The lower end of this range (2 m) corresponds
to the average size of relatively large objects such as plastic Sheets, large equipment,
and furniture that are typically in MSW. The higher end of this range (10 m)
corresponds to the typical width of the working face of a landﬁll. Based on the typical
thickness of waste lift, we used Ky equal to 0.5 m.

The variance in lognormally distributed hydraulic conductivities reﬂects the
spatial variation in the properties of waste. The presence of relatively impermeable
objects, such as plastic foils, increases variance as it reduces hydraulic conductivity
and enhances channeling.

We used the correlation lengths and log variance sets presented in Table 2 to
Simulate the effect of spatial variation in the hydraulic conductivity of waste on the
leachate ﬂux that can be recirculated at steady-state ﬂow condition (Q,) and on
leachate ﬂow patterns. In Table 2, k, and A; are the correlation lengths in horizontal
and vertical directions, respectively, and o 2 is the variance of lognormally distributed
kW. For each of these sets, 25 realizations were simulated using HYDRUS-ZD. Each
of these realizations had a unique hydraulic conductivity ﬁeld.

The leachate ﬂux at steady-state (Q,) was estimated for each realization using
HYDRUS-ZD. A steady-state condition was assumed to have reached when the total
ﬂux collected in the two LCS pipes equalized the total ﬂux injected in the three HTS.
At steady-state, Qs is independent of the initial water content of the model domain.

Following the simulation of each realization simulated by HYDRUS-ZD, MCA
was used to determine the recursive average of Q, (denoted as Q,) using Equation 4 as

follows,

62

Table 2. Summary of recursive Qs and standard deviation in Q5 for realizations using

heterogeneous and anisotropic waste properties.

 

 

 

Standard
Scaling Factors and variance to simulate Recursive Q, Deviation
heterogeneous and anisotropic waste (mJ/m/d) in Simulated Q,
(m3/m/d,
Setl:7»,=10 m,}\q=0.5 m, and02=0.42 3.12 1.27
Set 2: A,=2m,7\¢=0.5 m, and 62:0.42 3.44 1.14

 

 

 

 

63

'—1 Q3-
Qa- =17.an1'4'5 it (4)

l

 

where i is the iteration number or the realization count that ranges from 1 to N; N is

the number of realizations used to estimate Q, by MCS; Q is the recursive average of

Q, at the ith iteration; Q4.-. is the recursive average of Q8 at the (i-l)th iteration, and Q5,

is the leachate ﬂux for i“1 iteration.

Due to the signiﬁcant amount of time required for each Simulation (about 5-10
days for each Simulation), we limited total number of simulation (N) to 25 for each set
of geostatistical parameters listed in Table 2. More simulations would have reduced
the uncertainty in Q3. However, 25 simulations were enough to satisfy the key

objectives of this study.

RESULTS

Homogeneous and Isotropic Properties

First set of simulation was carried out where waste was Simulated as homogenous and
isotropic material having kW = 104 cm/S. A leachate injection pressure head, Hi, equal
to 2 m was used for the three horizontal trenches (Figure 1). Figure 3 presents the
leachate ﬂow distribution at steady-state for the case where waste is modeled as
homogenous and isotropic material. “Dry zones” or relatively low moisture content
zones between horizontal trenches can be observed in Figure 3. Dry zones could be

reduced by reducing the horizontal trench spacing or by applying a greater injection

head (Haydar and Khire 2005).

64

HT Spacing:
5" 20 m /—“Dry Zone”

 

 

 

z 90% Saturatio
Homogeneous and
Isotropic MSW
~ 5

 

Figure 3. Distribution of wetted area corresponding to degree of saturation > 90% at
steady-state using homogeneous and isotropic pr0perties of waste having kW = 10'4
cm/s and H,-= 2 m.

65

Heterogeneous and Anisotropic Properties

The next set of simulations were carried out where waste was simulated as a
heterogeneous and anisotropic material using realizations that correspond to the
correlation lengths and variances for three sets presented in Table 2. For each set in
Table 2, more than 25 realizations were Simulated. The reference hydraulic
conductivity of waste, kW, was set at 104 cm/s and the leachate injection pressure
head, H,-, was set at 2 m. For each set, the Simulated steady-state leachate ﬂux, Q,, for
each realization was converted into a recursive average, Qa, using Equation 4. Figure
4 presents the recursive average of leachate ﬂux at steady-state for each realization for
the two sets of correlation length.

After sorting the Simulated Q,, the ascending trend presented in Figure 4 for the
25 simulations of each set was sufﬁcient to deduce the effect of spatial variation on
Q,. When the correlation lengths are relatively short, the injected leachate does not
have to travel far before channeling or ﬁnding a zone having relatively large kW.
When the correlation lengths are relatively Short and the magnitude of variance is
relatively large, it simulates a waste matrix where waste pockets having relatively
high kW are interconnected. Injected leachate will avoid low kW pockets and follow
preferential ﬂow through high kW channels. Such interconnected pockets having
relatively high kW will lead to a Shorter ﬂow path for leachate resulting in greater
hydraulic gradient and, hence, greater steady-state leachate ﬂux. The area wetted by
the recirculated leachate for all Simulations in Set 2 is relatively narrow. When the
correlation lengths are relatively short, the risk of leachate break outs is minimal.

Table 2 presents the resulting Q, after 25 realizations and its standard deviation.

The standard deviation for Q3 obtained from Sets 1 and 2 decreased with decrease in

66

 

3.5

IIIIITTWITWrTIIIjTII

    

 

 

Recursive Leachate Flux, Q (maid/m)
N
0|

 

2 .
i . :
15 r Set1zkx=10m,kz=0.5m, a'=0.42j
1 I l 1 1 1
O 5 10 15 20 25

Number of realizations

Figure 4. Simulated recursive average leachate ﬂux versus number of realizations for
geostatistical realizations corresponding to Sets 1 and 2.

67

the correlation length. This occurs because a decrease in the correlation length causes
shortening of ﬂow path, thus reducing variation in the estimation of Q,. If more
realizations were conducted, it would reduce the standard deviation or uncertainty in
recursive Q,. However, it would have signiﬁcantly increased the computational time.
Simulation results presented in Table 2 indicate that, if horizontal correlation
length is increased, it reduces recursive Q,. This occurs because the increase in the
horizontal correlation length results in stratiﬁcation of the system into low and high
hydraulic conductivity layers. In such stratiﬁed system, when leachate ﬂow is
perpendicular to the layers, overall hydraulic conductivity is skewed towards the
lower hydraulic conductivity layers. Thus, stratiﬁcation lowers the leachate ﬂux when
leachate has to ﬂow perpendicular to the stratiﬁed layers. However, such stratiﬁcation
can create conduits made of zones having relatively high kW. This may lead to
leachate break outs from the Side slopes of a landﬁll if sufﬁcient buffer is not

maintained between the leachate injection point and the sides of the landﬁll.

Leachate Flow Patterns

Figure 3 presents the wetted area corresponding to greater than or equal to 90%
degree of saturation when the hydraulic properties were simulated as homogeneous
and isotropic. Figure 3 also shows that the wetted area for each HT is fairly uniform.
The maximum lateral spread of leachate from the leachate injection point (the center
of HT) is about 5 m for leachate injection head equal to 2 m.

Figure 5 presents the wetted areas at steady-state ﬂow condition for a typical
realization when the hydraulic properties of waste were Simulated with heterogeneity
and anisotropy using correlation lengths and variance corresponding to Sets 1 and 2.
Each realization yielded a different wetted area pattern. The key difference between

Set 1 and Set 2 is that Set 1 uses 2», equal to 10 m and Set 2 uses 2», equal to 2 m.

68

“Low Saturation

“Dry Zone” HT Pocket” Lateral Spreading
ii I 7 of Leachate

> 90% saturation
‘ i a u

 

 

‘

(*0 Mm; -

 

 

 

 

Leachate Channeling

“Low Saturation Pockets” r—“DI'Y Zone”

> 90% Saturation [$1
~ 8 m

Figure 5. Distribution of wetted area corresponding to degree of saturation z 90% for
geostatistical realizations corresponding to Set 1 (it, = 10 m, ll: 0. 5 m, and o 2" =
0. 42) (a); and Set 2 (3.,= 2 m, Aq= 0. 5 m, and o 2 = 0. 42) (b) for lognormal average
kW= 104cn1/sandH,= 2m.

 

 

69

Thus, it, for Set 1 is greater than it, for Set 2. Due to greater 70,, the maximum lateral
spread of the injected liquid for Set 1, which is equal to 15 m (Figure 5a) is greater
than the maximum lateral Spread of 8 m for Set 2 (Figure 5b). However, dry zones
and low saturation pockets were observed in both sets.

The non-uniform wetted area distribution demonstrates how heterogeneity and
anisotropy can lead to potential leachate break out(s) from the sides of a landﬁll. This
ﬁnding suggests that maintaining sufﬁcient buffer distance between the leachate
injection point and the side slopes of a landﬁll is critical for minimizing the potential
for leachate break outs. The leachate breakout distance was deﬁned as the distance
traveled by the injected leachate beyond an “imaginary” side slope of the landﬁll.
Figure 6a presents a schematic to illustrate the leachate break out distance. The side
Slope of the landﬁll was assumed equal to 1V:4H based on the typical side slopes
used for landﬁlls.

Figure 6b presents the leachate breakout distance measured for each of the
simulations for Sets 1 and 2. The leachate breakout distance ranged from 2 to 8 m as
shown in Figure 6b. Although, the leachate breakout distance was Similar for both
Sets of simulations, it was noted that the lateral travel of injected leachate for
simulations in Set 1 was relatively greater than that for Set 2. This is due to the
greater correlation length in Set 1 compared to Set 2. Figure 6b shows that buffer

distance shall be greater than 10 m to minimize leachate breakouts. Because 2. (71,, 71,,

0, 2,1) cannot be measured for a given waste matrix or MSW cell in the ﬁeld, the
correlation lengths listed in Table 2 were used to provide an initial estimate for what
buffer is appropriate to minimize leachate breakouts. Field data on leachate breakouts

and buffer distance needs to be collected to verify or estimate correlation lengths.

70

HT

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Leachate Breakout (a)
Distance
""""""" IP'TIP'
‘ ,,,,,, l 1 unis:
4 <>
~Waste~
Wetted Area
10 l r ‘ v 7 ‘ I
A I Seti 7 (b) ,
E 1 D Set2 ‘
"’ .'
o1 -.. a
0 8 g
C .
”3 i I I 1:1 I i
a . I :1 1:1 I
...- 6 LL .................................................................................... a
a - I 13
o P
x f- . D I I I3 - C1 G3
I! 5
' I 1:1 1:1 I I I 1: :1 I [:1
g 4 L. ............................ ................................................ _.
w l I Cl 1:1 I I 1:1
e-0 L i
«I . 1:1 1:1 1:1 I I I I I
.5 l
n ._ ................ I .................... I ........................... E1 ............
o 2 : I
..l |
i- -l
0 l I . 1 - i . 1 . . i 1 1 1 . 1 1
0 5 10 15 20 25

Realization Number

Figure 6. Schematic of leachate breakout distance (a); and Simulated leachate
breakout distance for simulations using Sets 1 and 2 for kW = 10'4 curls and H,- = 2 m

(b)-

71

Hence, in order to reduce leachate breakouts from the sides of a landﬁll, we
recommend these design and operational guidelines: (1) buffer distance between the
leachate injection point and the sides of a landﬁll Shall be greater than 15 m. If
leachate injection pressure is greater than 2 m, further increase in buffer is necessary;
(2) daily covers consisting of relatively low hydraulic conductivity soil shall be
avoided; (3) daily cover layers made up of soils shall be scraped off to make the daily
cover layer not continuous for greater than 10 m distance; and (4) hydraulically
impermeable plastic sheets or tarps used as daily covers Shall be removed prior to
placing the next lift of waste. We also recommend design engineers to use the results
presented in this study as a guideline. However, site-Speciﬁc data should be used to

optimize the buffer distance and other design issues related to LRS.

SUMMARY AND CONCLUSIONS

In this numerical study, the effect of heterogeneity and anisotropy in the hydraulic
properties of waste on the magnitude of leachate ﬂux recirculated using a LRS
consisting of horizontal trenches was evaluated. Over 50 realizations were simulated
using HYDRUS-ZD to evaluate the leachate ﬂux as a function of heterogeneity and
anisotropy expressed by three sets of horizontal and vertical correlation lengths and
variance in the hydraulic properties of waste. The MCA was used to evaluate the
effect of heterogeneity and anisotropy.

Increasing horizontal correlation length resulted in more lateral Spreading of
leachate and relative decrease in leachate ﬂux. The magnitude and the pattern of
wetted area was greatly affected when heterogeneity and anisotropy of waste was
considered. Introduction of heterogeneity and anisotropy resulted in non-unifonn

leachate wetting. Increase in the horizontal correlation length resulted in greater

72

wetting in the horizontal direction with greater potential for leachate breakouts. A
minimum buffer distance of 15 m between the side slopes of landﬁll and the leachate
injection point is needed to minimize leachate breakouts assuming the horizontal

correlation length is less than or equal to 15 m.

ACKNOWLEDGEMENTS

We sincerely appreciate the help provided by Dr. Jirka Simunek related to running the
computer model HYDRUS-ZD and Dr. Shu-Guang Li’s assistance in the selection
and interpretation of geostatistical parameters. However, the results and opinions

presented in this manuscript are those of the authors.

73

REFERENCES

Bendz, D. (1998). “Generation of Leachate and the Flow Regime in Landﬁlls.”
Report No 1023, Department of Water Resources Engineering, Lund University,
Sweden.

Bleiker, D., F arquhar G., and McBean, E. (1995). “Landﬁll Settlement and the Impact
on Site Capacity and Refuse Hydraulic Conductivity.” Waste Management and
Research, Vol. 13, 533-554.

Bleiker, D., McBean, E., and Farquhar, G. (1993). “Refuse Sampling and
Permeability Testing at the Brock West and Keele Valley Landﬁlls.”
Proceedings of the Sixteenth International Madison Waste Conference, 548-567.

Doran, F. (1999). “Lay Leachate Lay.” Waste Age, April Issue, 74-79.

Fellner, J., Huber, R., Dbberl, G., and Brunner, P. (2003). “Hydraulics of MSW
Landﬁlls and its Implications for Water Flow Modeling.” Proceedings of Ninth
lntemational Waste Management and Landﬁll Symposium, Sardinia, Italy,
October.

Fenton, G. and Grifﬁths, D. (1996). “Statistics of Free Surface Flow Through a
Stochastic Earth Dam.” Journal of Geotechnical Engineering, ASCE, Vol. 12,
No. 6, 427-436.

Fungaroli, A. and Steiner, R. (1979). “Investigation of sanitary landﬁll behavior.”
Vol. 1, Final Report, US. EPA 600/2-79-053a.

Haydar, M. and Khire, M. (2005). “Leachate Recirculation Using Horizontal
Trenches in Bioreactor Landﬁlls.” Journal of Geotechnical and
Geoenvironmental Engineering Journal, ASCE, July.

Hughes, G., Landon, R., and Farrolden, R. (1971). “Hydrogeology of Waste Disposal
Sites in Northeastem Illinois.” EPA Solid Waste Management Series, SW-12d.

Khire, M. and Haydar, M. (2005). “Leachate Recirculation Using Geocomposite
Drainage Layer in Engineered MSW Landﬁlls.” Proceedings of
GeoFrontiersZOOS, ASCE, Austin, TX, 23-26 Jan.

Khire, M. and Haydar, M. (2003). “Numerical Evaluation of Granular Blankets for
Leachate Recirculation in MSW Landﬁlls.” Proceedings of the Ninth Sardinia
Solid Waste Conference, Cagliary, Italy, October.

Korﬁatis, G., Demetracopoulos, A., Bourodimos, E., and Nawy, E. (1984). “Moisture

transport in a Solid Waste Column.” Journal of Environmental Engineering,
Vol. 110, No. 4, 789-796.

74

Maier T. and Vasuki, N. (1996). “Expected Beneﬁts of a Full-scale Bioreactor
Landﬁll.” Proceedings of WASTECON, Portland, USA, 179-185.

McCreanor, P. and Reinhart, D. (1998). “Hydrodynamic Modeling of Leachate
Recirculating Landﬁlls.” Proceedings of Swedish Landﬁll Symposium,
Bioreactor Technology Session, October.

Mehta, R., Barlaz, M., Yazdani, R., Augenstein, D., Bryars, M., and Sinderson, L.
(2002). “Refuse Decomposition in the Presence and Absence of Leachate
Recirculation.” Journal of Environmental Engineering, Vol. 128, No. 3, 228-
236.

Miller, D. and Emge, S. (1997). “Enhancing Landﬁll Leachate Recirculation System
Performance.” Practice Periodical of Hazardous, Toxic, and Radioactive Waste
Management, July, 1 13-119.

Miller, E. and Miller, R. ( 1956). “Physical theory for Capillary Flow Phenomena.”
Journal of Applied Physics, Vol. 27, 324-332.

Mualem, Y. (1976). “A new model for predicting the hydraulic conductivity of
unsaturated porous media.” Water Resources Research, Vol. 12, 513-522.

Oweis, 1., Smith, D., Ellwood, R., and Greene, D. (1990). “Hydraulic Characteristic
of Municipal Refuse.” Journal of Geotechnical Engineering, Vol. 116, No. 4,
539-553.

Powrie, W. and Beaven, R. (1999). “Hydraulic Properties of Household Waste and
Implications for Liquid Flow in Landﬁlls.” Proceedings of the Institution of
Civil Engineers, Geotechnical Engineering Association, October.

Qian, X., Koemer, R., and Gray, D. (2002). Geotechnical Aspects of Landﬁll Design
and Construction, Prentice Hall, New Jersey.

Reinhart, D. and Carson, D. (1993). “Experiences with Full-scale Application of
Landﬁll Bioreactor Technology.” The 31St Annual Solid Waste Exposition of
the Solid Waste Association of North America, San Jose, California, August.

Richards, L. (1931). “Capillary Conduction of Liquids in Porous Medium.” Journal of
Physics, 318-333.

Rosqvist, H. and Destouni, G. (2000). “Solute Transport through Preferential
Pathways in Municipal Solid Waste.” Journal of Contaminant Hydrology, Vol.
46, 39-60.

Simunek, J ., Sejna M., and Van Genuchten, M. Th. (1999). HYDRUS 2D, Simulating
water ﬂow, heat, and solute transport in two-dimensional variably saturated
media, Version 2.0, US Salinity Laboratory, ARS/USDA, Riverside, California
and lntemational Ground Water Modeling Center, IGWMC- TPS 53, Colorado
School of Mines, Golden, Colorado.

75

SWANA (2002). “The Solid Waste Manager’s Guide to the Bioreactor Landﬁll.”
SWANA, October.

van Genuchten, M.Th. (1980). “A Closed-fonn Equation for Predicting the Hydraulic
Conductivity of Unsaturated Soils.” Soil Science Society of America Journal,
Vol. 44, 892-898.

Vogel, T., Cislerova, M., and Hopmass, J. (1991). “Porous Media with Linearly
Variable Hydraulic Properties” Water Resources Research, Vol. 27, No. 10,
2735-2741.

Zeiss, C. and Uguccioni, M. (1995). “Mechanisms and Patterns of Leachate Flow in
Municipal Solid Waste Landﬁlls.” Journal of Environmental Systems, Vol. 23,
No. 3, 247-270.

Zomberg, J., Jemigan, B., Sanglerat, T., and Cooley, B. (1999). “Retention of Free

Liquids in Landﬁlls Undergoing Vertical Expansion.” Journal of Geotechnical
and Geoenvironmental Engineering, ASCE, Vol. 125, No. 7, 583-594.

76

PAPER NO. 3: NUMERICAL EVALUATION OF PERMEABLE
BLANKETS FOR LEACHATE RECIRCULATION IN MSW
LANDFILLS

ABSTRACT

Leachate recirculation in municipal solid waste (MSW) landﬁlls is commonly
achieved by using horizontal trenches and/or vertical wells. This paper presents a
numerical modeling study of a leachate recirculation alternative using permeable
blanket. The performance of permeable blanket was evaluated by comparing
recirculated leachate ﬂux under steady-state conditions to that for horizontal trench.
The ﬁnite-element saturated/unsaturated ﬂow model HYDRUS-2D was used for the
numerical evaluation. The effect of hydraulic properties of permeable blanket material
on the recirculated leachate ﬂux was also evaluated. The numerical modeling results
indicate that signiﬁcantly higher leachate ﬂux can be recirculated using permeable
blankets compared to an equivalent system of horizontal trenches spaced at 10m

center to center. Field—scale testing of permeable blankets alternative is in progress.

INTRODUCTION

Recirculation of leachate in MSW landﬁlls offers many beneﬁts (Maier and Vasuki
1997), including: (1) reduction in leachate treatment and disposal costs; (2)
acceleration and increase in waste settlement resulting in creation of airspace; (3)
acceleration in the generation of landﬁll gas and hence potential for higher short-tenn
revenues from gas-to-energy system (Mehta et al. 2002); and (4) reduction in the risk
associated with contamination during off-site transportation, treatment, and disposal

of leachate.

77

The most commonly used leachate recirculation system (LRS) in MSW landﬁlls
consists of horizontal trenches (HT) ﬁlled with high conductivity material with a
perforated pipe at the center. Leachate is recirculated by injecting leachate at positive
pressures ranging from 0 to 5 m. Typical dimensions of a horizontal trench are 0.6m
wide by 1m deep (Figure 1). A horizontal trench may be the most commonly used
LRS, however, it is not the most efficient method to recirculate leachate. Horizontal
trench typically creates “dry pockets” between adjacent horizontal trenches where
recirculated leachate may not reach [Figure 2(a)]. Such dry pockets increase
differential settlement of waste and reduce the quantity and the rate of landﬁll gas
generation. Differential settlement potentially increases the cost of maintaining the
ﬁnal cap.

The permeable blanket (PB) alternative evaluated in this paper consists of a 0.1
to 0.15m thick layer of relatively high hydraulic conductivity material laid on waste
surface with a perforated pipe running through the center of the layer [Figure l(b)].

A permeable blanket can be made of coarse sand, gravel, or permeable recycled
materials such as, crushed recycled glass or Shredded tires. The widths of permeable
blankets evaluated in this paper ranges from 30 to 60 m. Permeable blankets offer
these advantages over conventional horizontal trenches: (1) signiﬁcant increase in
quantity of leachate that can be recirculated per unit mass of waste; (2) can achieve
relatively uniform distribution of moisture (or leachate) reducing dry pockets and
hence reducing differential settlement; (3) increased gas production rate; and (4) can
also be used for extracting landﬁll gas. A schematic of simulated water content
distributions and resulting dry pockets for an LRS consisting of horizontal trenches is
presented in Figure 2(a). A schematic of Simulated water content distributions for a

LRS consisting of a permeable blanket is presented in Figure 2(b). Figure 2(a) and

78

2(b) represent progressive snap shots of water content as the system reaches steady-
state. The material properties were assumed homogeneous and isotropic for waste and

other components.

NUMERICAL MODELING

HYDRUS-2D Computer Model

HYDRUS-2D is a computer model that can Simulate water, heat, and solute
movement in unsaturated, partially saturated, or fully saturated porous media
(Simunek et al. 1999). The program numerically solves the Richards’ Equation for

saturated-unsaturated water ﬂow. A 2-D form of Richards’ equation can be expressed

as follows:
66 6k
— = —V. k .V +— — S 1
a: [ (W) 1.1/1 62 ( )
where, 0 = volumetric water content; w = matric suction head; k = hydraulic

conductivity, which is k, for saturated soil but is strongly dependant on the soil
suction; z = vertical dimension; S = volume of water removed per unit time per unit
volume of soil by plant water uptake (sink term); and t = time.

The model uses van-Genuchten function for soil-water characteristic curves and
van-Genuchten-Mualem model for predicting the unsaturated hydraulic conductivity
function. The governing ﬂow and transport equations are numerically solved using
Galerkin-type linear ﬁnite-element schemes. Depending upon the scale of the problem
domain, the matrix equations resulting from the discretization of the governing
equations are solved using either Gaussian elimination for banded matrices, a
conjugate gradient method for symmetric matrices, or the ORTHOMIN method for

asymmetric matrices.

79

donut _Snoutog 8m mace 3 @8333 Eamon—Hum so.“ :85 3V “aﬁmDMQDE memm: 033883 gov—San 0385a
use scene 5555: 2285:0280 a mean 5:62 >52 5 20.525308 826mm: mo eouﬂaﬁmm com _oeoE 3:38:00 A8 A PEME

        
  

 

E2m>m 8:02.00 oumcomoh u mo; .0
co=m_:2_omm 2938.. u m; .m
Excﬂm mEmmEon. n ma .v
.9552 5ch _ScoEori 1 PI .m
ommEEQ w._<om OH ._.Oz .N
@5859”. .862
E 5:95:
“9.52m mime—Eon. em \loamEED oEmeoa
E

 

 

ALI .2858 new:

E.
28.6.88 85 S emanated

Em» d S s .e |+ Abmuczom
mom”. ommqoowv

moi moi.
Ammtmocsom
x2“. 96S 55.. W e r

 

ALI 58:38
281 ESmcoov
SE S Basted

E

  
   

       

25:25am WW :38 near:

89 mo WW<W> ..
/., .. .. .
ayagwx deﬂate 1%....

\

               

wormuczom x2... oLoN

80

48:55 03885.“ a SV 95 account _Snouton m 3 magma”. 88m? noun—3&8“ 8382 How nounpmummw 23602 .N 8sz

_‘

mvd

..yérxlm... um 31am .515;
_ _ W _ _ . _ _ 4

ovd mmd omd nNd omd 26 26 mod cod

5:880 SE3 oEoEEO>

 

81

Conceptual Model and HYDRUS-2D Input

The conceptual model used for the simulation of moisture dynamics for LRS consisting
of a horizontal trench and permeable blanket is presented in Figure 1. Figure 1(c)
presents the detail for the horizontal trench and Figure l(b) presents the details for the
permeable blanket. Leachate was simulated as pure water and the effect of biochemical
reactions that occur within a landﬁll on the hydraulics of leachate recirculation was
ignored. Waste and other materials were assumed to be porus, homogeneous, and

isotropic.

Boundary Conditions and Mass Balance

All external boundaries were simulated as zero ﬂux boundaries {Figure 1(a)]. Leachate
movement as a result of percolation from the cap or waste above model domain was not
simulated. The perforated pipe used for leachate injection was simulated as a constant
head boundary. The simulated leachate injection heads ranged from zero (gravity
application) to 5 m. The leachate injection head assigned as the constant head is exclusive
of head loss in pipes, joints, manifolds, and pumps used in a typical LRS. Leachate
collection pipes embedded in the leachate collection system (LCS) were simulated as
seepage face boundaries. For a seepage face boundary, the hydraulic pressure head is
always maintained zero or less. The diameter of such LCS pipes was assumed equal to
0.15 m.

The minimum size of the ﬁnite-elements used for discretization of the problem
domain, the time step, and the error tolerances for pressure head and water content were

selected such that cumulative water balance error does not exceed 0.1%.

82

Hydraulic Properties

Total three material types were used for simulating horizontal trench and permeable
blanket LRS. These three materials were simulated as homogeneous and isotropic soils.
The saturated hydraulic conductivities of these materials are listed as follows in the
parentheses.

0 Solid waste (saturated hydraulic conductivity, kw = 10*6 to 10'5 m/s)

0 horizontal trench or permeable blanket drainage material (km or k3 = 10'3 to 10'2

m/s)

0 LCS drainage material (kLCS = 10'2 m/s)
The unsaturated hydraulic properties for the materials listed above consisted of van
Genuchten ﬁtting parameters for the soil-water characteristic curves. The unsatuted
hydraulic properties were selected from the database of HYDRUS-2D for soils having

saturated hydraulic conductivities close to the assumed hydraulic conductivities.

Geometry and Dimensions

The dimensions of the simulated horizontal trench were 0.6 m wide by 1 m deep. The
thickness of the simulated permeable blanket was 0.15 m. The width of the permeable
blanket ranged from 30 to 60 m. The thickness of the LCS layer, ths, was assumed equal
to 0.3 m. The spacing between the adjacent leachate collection pipes, d, was assumed
equal to 60 m. The slope (s) of the LCS was assumed equal to 3%. The LCS design
parameters (ths, kLcs, d, and s) were selected such that the maximum leachate pressure
head of leachate on the liner did not exceed the US. Subtitle D requirement of 0.3 m.
The vertical distance (D) between the horizontal trench or permeable blanket and the top

of the LCS was assumed equal to 15 m. Distance from top of the horizontal trench or

83

permeable blanket to the upper zero ﬂux boundary was assumed equal to 10 m to contain

all leachate movement under the leachate injection heads not exceeding 5 m.

SIMULATION RESULTS

HYDRUS-2D simulation output was analyzed to compare performance of permeable
blanket to horizontal trench. The steady-state recirculated leachate flux (hereafter,
referred to as leachate ﬂux) was used as the primary criterion for the comparison. In
addition, the effect of saturated hydraulic conductivity and the width of permeable

blanket on leachate ﬂux was also evaluated.

Horizontal Trench

The plot of leachate ﬂux for a single horizontal trench as a function of leachate injection
pressure head (hereafter, referred to as injection head) is presented in Figure 3. The
leachate flux at injection head equal to zero for kw equal to 106 m/s is about 0.3 m3/d per
meter length of the LR pipe perpendicular to the plane of the paper. Leachate ﬂux
increases with the increase in injection head. This relationship is curvilinear. The effect
of kw on leachate ﬂux is also presented in Figure 3. The simulation results indicate that
the leachate flux is directly proportional to kW. Thus, if kw is increased or decreased by an
order of magnitude, assuming all other parameters constant, the leachate ﬂux would also

increase or decrease by an order of magnitude.

Permeable Blanket

The simulation results for LRS using permeable blanket and comparison of the results

with horizontal trench are presented in Figures 4 to 6. All permeable blanket simulations

84

 

100 _ .............. .1 ................ I ............... i ...............

_x
O

8

Leachate Flux, Q (maldlm)

 

 

 

Horizontal Trench

 

O 2 4 6 8 1O
Leachate Injection Pressure Head, H, (m)

Figure 3. Simulated steady state recirculated leachate ﬂux versus leachate injection
pressure head for single horizontal trench for kw = 10‘, 5x10'6, and 10-5 m/s.

85

were conducted for kw equal to 104’ m/s. Unless speciﬁed differently, kg was assumed

equal to 10'2 m/s and width of permeable blanket was assumed equal to 60 m.

Effect of Injection Head

Similar to the horizontal trench, as the injection head is increased, the leachate ﬂux
increased for the permeable blanket. However, the increase is log-linear for injection
heads greater than the breakthrough pressure head (Figure 4). The breakthrough pressure
head is the minimum injection head (greater than zero) required for leachate to
breakthrough or flow across the entire width of the permeable blanket. Theoretically, for
a perfectly horizontal permeable blanket, an injection head slightly greater than zero will
cause breakthrough across the width of the blanket. At injection head equal to 1 m and 5
m, the simulated leachate ﬂux for a 60-m-wide permeable blanket is 6.5 and 8.5 m3/d per
meter length of the blanket.

At injection head equal to zero, permeable blanket behaves similar to a horizontal
trench. At injection head equal to zero, the saturated region at steady-state is narrow at
the top and widens for the lower portion. At injection heads greater than zero, the width
of the saturated region is constant equal to the width of the permeable blanket [refer to

Figure 2(b)].

Effect of Width of Permeable Blanket

The leachate ﬂux increases linearly with increase in the width of the permeable blanket.
Figure 4 presents the leachate flux as a function of injection head for permeable blankets
having 30 m and 60 m widths. At injection heads greater than 0m, the leachate ﬂux for

60-m-wide permeable blanket is exactly two times the leachate ﬂux for 30-m-wide

86

 

 

 

60m-wide Permeable Blanket ----------------- -
............................. : """"'----------------..

 

 

-..- .---.---.----.lc.---.-..o-..--..-...-...a.o..-o-.-.---.---o--c-.—

   

------------ 30m-wide Permeable Blanket

  
   

Breakthrough Pressure Head

-------------------------------

..........................................................................

.............

...........

  

----------.------ ..........

1 Single Horizontal Trench
§.(0 6m wide x 1 m deep)

Leachate Flux, Q (maid/m)

 

 

 

0.1 i i i i
o t 2 3 4 5

Leachate Injection Pressure Head, H, (m)

Figure 4. Simulated steady-state recirculated leachate flux versus leachate injection
pressure head for single horizontal trench and for 30-m-wide and 60-m-wide permeable
blankets for kw- " 10 m/s and kHT& k3 = 102 m/s.

87

permeable blanket at equal injection heads. The authors believe that if the width of the
permeable blanket is reduced below 30 m, the leachate ﬂux will continue to reduce
linearly until the width is small enough for the permeable blanket to behave similar to a
horizontal trench. At that point, the linear relationship will not hold. The authors have not
completed these simulations to evaluate the smallest width of permeable blanket at which

it would behave similar to a horizontal trench.

Effect of Hydraulic Conductivity of Permeable Blanket

Figure 5 presents the leachate flux as a function of injection head for the saturated
hydraulic conductivities of the permeable blanket, k3, equal to 103, 5x103, and 10'2 m/s.
Hydraulic conductivities of gravel and sand range from 10’3 to 10’2 m/s and 10'5 to 10'3
m/s, respectively. The leachate ﬂux increases as Its increases. However, once k3 is equal
to or greater than 5x10'3 m/s or when the injection head is equal to or greater than 1.5 m,
the increase is negligible. Simulation results presented in Figure 5 indicate that if k3 is
equal to 10'3 m/s, the injection head shall be greater than or equal to l.5 m to achieve

leachate ﬂux similar in magnitude to that for k3 equal to or greater than 5x10'3 m/s.

Permeable Blanket and Horizontal Trench Equivalency

Figure 5 indicates that the amount of leachate ﬂux that can be recirculated using a 60-m-
wide permeable blanket is about one order magnitude higher than a single horizontal
trench at zero injection head and about 6 times higher at 5 m injection head. In order to
evaluate how many single trenches, if used simultaneously, would result in leachate ﬂux
similar in magnitude to a 60-m-wide permeable blanket, seven trenches spaced at 10 m

center to center were simulated. The spacing between the farthest trenches was

88

 

 
   
 

  

 

 

 

é . .
2 ...................................................................
n
E, ..........................................................................
O“D ' _3' g g
g , .......... KB=5X10 W3 ......3 ............. E ............
E g s 3 g i _
8 ‘\ """""" """"""" g """""" ‘3
g E K =10'3m/s 3 : : .
o ........‘-.--.B. ........................................................... _‘
N ' .
w -
-| :
"""""""" a """""""" : """""" 1
0 i l 60In-wide Permeable Blanket 3
l l l
O 0.5 1 1.5 2 2.5 3

Leachate Injection Pressure Head, H, (m)

Figure 5. Simulated steady-state recirculated leachate ﬂux versus leachate injection
pressure head for 60-m-wide permeable blankets having k3 = 103, 5x10'3, and 10'2 m/s.

89

maintained at 60 m — same as the width of the permeable blanket. The total leachate ﬂux
from the seven horizontal trenches is presented in Figure 6. The leachate ﬂux that can be
recirculated using a 60m-wide permeable blanket exceeds the total ﬂux from seven
horizontal trenches at all injection heads ranging from O to 5m. The difference reduces as

the injection head increases.

Effect of Settlement of Permeable Blanket

The total settlement of MSW in landﬁlls ranges from 10% to 30% of initial thickness
(Sowers 1972). The composition of waste, climate, leachate recirculation, and other
physical and biochemical factors impact the settlement. Due to settlement of waste,
permeable blankets constructed within waste will also undergo settlement. The effect of
settlement of waste on the leachate flow through a 60-m-wide permeable blanket was
evaluated by simulating a 3-m-deep sag at the center of the permeable blanket. The 3-m-
deep sag simulates a 3 m differential settlement of waste. The simulation results indicate
that for leachate injection head less than the magnitude of sag (3 m in this case), the
leachate ﬂux is less than that for a perfectly horizontal permeable blanket (zero sag).
However, for injection heads greater than or equal to the sag, the leachate ﬂux is equal to

the leachate ﬂux for a perfectly horizontal permeable blanket.

SUMMARY AND CON CULSION S

This study presents a numerical evaluation of permeable blankets — an alternative
developed by the authors for leachate recirculation in MSW landﬁlls. The steady-state
leachate flux that can be recirculated at injection heads ranging from O to 5 m was

simulated using the computer program HYDRUS-2D for the conventional horizontal

9O

 

 

 

 

 

10 l ! 1 !
: 60m-wide Permeable Blanket .
E 8 — -------------- ------------------------ 3
E i s i
a I
E, ~ == ~
on 6 _.-_.. """"""""""""""" “
x“ _ . . . . _
2 " I "-3 E i E -
u. 4 — ------------ i- --------------- -------------- -
q. _ :: X : : : _
*5 3 7 Trenches Spaced at 10m (.10
5 (each trench, 0. 6m wide x 1m deep)
I“ 5 1 :
m 2:: ----- . ' ----------------------------- ——
_. Single Trench E ‘5:
E g -
0': i i 1 _
O 1 2 3 4 5

Leachate Injection Pressure Head, H, (m)

Figure 6. Simulated steady-state recirculated leachate flux versus leachate injection
pressure head for single horizontal trench, seven horizontal trenches spaced at 10m c/c,
and a 60-m-wide permeable blanket having kg and km = lO'2 m/s.

91

trench and permeable blanket. Based on the numerical evaluation, a 60m-wide
permeable blanket can recirculate greater amount of leachate ﬂux at injection heads less
than 5m, compared to total ﬂux in seven horizontal trenches spaced at 10 m center to
center.

The capital cost for the construction of a 60-m-wide permeable blanket is expected
to be less than that for seven horizontal trenches for most sites. However, economical
gains shall be evaluated on a site—speciﬁc basis. A permeable blanket distributes leachate
more evenly compared to a set of horizontal trenches. Thus, dry pockets and hence
differential settlement of waste is expected to be smaller for landﬁlls where LR is done
using permeable blankets. The summary of the key results is as follows:

As the injection head is increased, the leachate ﬂux increases for both horizontal
trench and permeable blanket. The relationship is curvilinear for horizontal trench. The
relationship is log-linear for permeable blanket for injection heads greater than the
breakthrough pressure head. Permeable blankets can recirculate relatively larger leachate
ﬂux at lower injection pressures. Thus, a pump with relatively small head but high ﬂow
shall be necessary.

The leachate ﬂux linearly increases with increase in the width of permeable blanket.

The increase in the hydraulic conductivity of permeable blanket material results in
increase in leachate flux. However, for hydraulic conductivities greater than or equal to
5x10'3 m/s, the difference is negligible.

Effect of sag in permeable blanket as a result of settlement of waste indicated

reduction in leachate flux for injection heads less than the magnitude of the sag. For

92

injections heads greater than the magnitude of sag, the leachate ﬂux is similar to a
perfectly horizontal permeable blanket.
Field veriﬁcation of performance of permeable blankets must be conducted to verify

the numerical modeling results presented in this study.

ACKNOWLEDGEMENTS

The authors are thankful to Dr. Jirka Simunek of US. Salinity Laboratory, US.
Department of Agriculture, for his input related to running the computer model
HYDRUS-2D. However, the results and opinions presented in this manuscript are those

of the authors and have not been reviewed by anyone else.

93

REFERENCES

Maier, TB, and Vasuki, NC. (1997). “Expected Beneﬁts of a Full-Scale Bioreactor
Landﬁll. A Compilation of Current Readings: Leachate Recirculation & Landﬁll
Bioreactors,” SWANA, July 1997.

McCreanor, RT. and Reinhart, DR. (2002). “Hydrodynamic Modeling of Leachate
Recirculating Landﬁlls,” Swedish Landﬁll Symposium, Oct. 1998.

Mehta, R., Barlaz, M., Yazdani, R., Augenstein, D., Bryars, M., and Sinderson, L. (2002)
“Refuse Decomposition in the Presence and Absence of Leachate Recirculation,”
Journal of Environmental Engineering, Vol. 128, No. 3, 228—236.

Simunek, J ., Sejna, M., and M. Th. Van Genuchten (1999). “The HYDRUS-2D Soﬁware
Package for Simulating the 2-D Movement of Water, Heat, and Multiple Solutes in
Variable Saturated Media, Version 2.0. US. Salinity Laboratory,” Agriculture
Research Service, USDA, Riverside, California.

Sowers, G. F. (1972). “Settlement of Waste Disposal F ills,” Proceedings of 8th

International Conference on Soil Mechanics and Foundation Engineering,
Moscow, 1972,207-210

94

PAPER NO. 4: LEACHATE RECIRCULATION USING
HORIZONTAL PERMEABLE BLANKETS IN BIOREACTOR
LANDFILLS

ABSTRACT

Conventional subsurface leachate recirculation or liquid injection methods for municipal
solid waste (MSW) landﬁlls are horizontal trenches and vertical wells. In this study, we
have presented numerical modeling and ﬁeld-scale testing of a recently developed
leachate recirculation system (LRS) called permeable blankets (PBS). HYDRUS-2D, a
saturated/unsaturated water ﬂow model was used to simulate the travel of injected
leachate in PBs. These parameters were numerically evaluated in this study: (1) hydraulic
properties of PB and waste; (2) geometry of PB; (3) settlement of PB; (4) leachate dosing
frequency; and (5) initial degrees of saturation of PB and waste. In the ﬁeld, at an MSW
landﬁll located in Michigan, the migration of injected leachate in a 60-m-wide by 9-m-
long by 0.15-m-deep blanket made up of crushed glass was measured using an automated
sensing system consisting of sensors embedded in the blanket. Leachate injection rates
simulated in this study and used in the ﬁeld ranged from 1.1 m3/hr to 3.6 m3/hr per meter
length of the injection pipe embedded in the PB. The key ﬁndings of the study are: (l) the
rate and maximum distance of travel of injected leachate are a strong function of the
hydraulic properties of the PB and underlying waste and the rate and frequency of
leachate injection; and (2) the maximum pressure head in the blanket due to liquid
injection does not exceed the injection pressure. The numerical modeling results and the
ﬁeld data indicated that PBs can be designed to inject liquids or recirculate leachate in

MSW landﬁlls. Long-term performance of such blankets needs to be evaluated.

BACKGROUND
Environmental and economical beneﬁts of leachate recirculation for municipal solid

waste (MSW) landﬁlls are well documented (Reinhart and Al-Yousﬁ 1996; Reinhart and

95

Townsend 1998; Pohland and Kim 1999). These beneﬁts include: (1) a reduction in the
leachate treatment and disposal costs; (2) accelerated decomposition and settlement of
waste resulting in an airspace gain; (3) an increase in the rate of gas production; (4)
improvement in the quality of recirculated leachate; and (5) potential reduction in the
post-closure care period and associated maintenance costs. Risks and drawbacks of
leachate recirculation include: (1) potential decrease in the factor of safety for slope
stability of landﬁlls; (2) potential increase in the leachate head on the liner if leachate
collection system is not designed to efﬁciently drain injected leachate; (3) potential
ﬂooding of the gas collection system; and (4) leachate seeps from the landﬁll side slopes
if adequate buffer distance is not provided. Hence, before implementation, landﬁll
operators should weigh the advantages and risks associated with leachate recirculation on
a site-speciﬁc basis (Haydar and Khire 2005a).

Leachate recirculation techniques can be broadly divided into surface and
subsurface application. Surface application consists of: (1) direct application of leachate
or spray irrigation of leachate on the landﬁll surface; and (2) surface ponding of leachate.
Climate dependency, odor problems, interference with daily operations, poor aesthetics,
and potential runoff of applied leachate into storm water management system are the
drawbacks of these surface application techniques. Conventional subsurface application
methods are: (1) vertical wells; and (2) horizontal trenches (Koemer 2000, Qian et al.
2002)

Vertical wells and horizontal trenches (HTS) are the most commonly used leachate
recirculation methods for relatively old and relatively new, MSW landﬁlls, respectively

(Maier and Vasuki 1997; Miller and Emge 1997; Haydar and Khire 2005a).

96

Disadvantages and limitations of vertical wells and horizontal trenches include: (1)
during installation, excavation of waste may cause odor problems; (2) relatively high
capital cost of construction; and (3) formation of “dry zones” if spaced too far apart
where injected leachate cannot reach and hence cannot wet the waste. Such dry zones can
lead to differential settlement of waste and hence can result in greater landﬁll cap
maintenance costs. Design guidelines for LRS consisting of horizontal trenches are
presented by Haydar and Khire (2005a). Haydar and Khire (2005a) suggested that HTs
should be spaced at 10 m or less spacing for injection heads 5 5 m to ensure a relatively
uniform wetting of the waste. Vertical wells can interfere with landﬁll daily operations.
However, key advantages of using vertical wells LRS is that such wells can be installed
after a landﬁll is capped or for retroﬁt landﬁlls. It is moSt cost effective to install

horizontal trenches before the landﬁll is capped at the design elevation in the landﬁll cell.

LEACAHTE RECIRCULATION USING PERMEABLE BLANKETS

Design and Construction Aspects

Permeable blankets are constructed by laying a relatively thin layer of permeable material
having relatively high hydraulic conductivity on a horizontal or inclined waste surface in
a landﬁll. Geotextiles placed directly above and below the blanket to separate the PB
from the surrounding porous material (e.g., soil, waste, etc.) and to prevent clogging of
the blanket. The thickness of such blankets can vary depending upon the material used
(e.g., shredded tires, pea gravel, crushed glass, GDL, etc.) and site-speciﬁc design and
operational variables. A perforated pipe is embedded in the blanket in the direction
parallel to the shorter or longer plan view dimension of the blanket where leachate is

injected under a positive pressure. The relatively high hydraulic conductivity of the PB

97

allows preferential travel of injected leachate or liquids within the blanket and wetting of
the underlying waste as the injected leachate inﬁltrates. The aerial dimensions and the
shape of the PB can vary depending upon the leachate recirculation needs, shape of the
landﬁll cell, relative contrast in the hydraulic conductivities of the blanket and underlying
waste, and leachate injection pressure or leachate injection rates.

Khire and Haydar (2003) conducted preliminary numerical modeling to compare
the hydraulic performance of HTs versus PBs. A constant injection pressure head was
used to simulate leachate ﬂux under steady-state condition. This ﬂux and the wetted areas
were used to evaluate the hydraulic performance of the simulated LRSS. The results
demonstrated that PBs are hydraulically more efﬁcient compared to HTs to unifome
wet the underlying waste. Khire and Haydar (2005) presented data from an instrumented
ﬁeld—scale PB made up of geocomposite drainage layer (GDL). Moisture content sensors
embedded just beneath the GDL blanket indicated that the injected leachate traveled at a
rate ranging from 8 to l8 m/hr for leachate injection rates ranging from 0.9 to 2.6

m3/hr/m. Long-term monitoring of the performance of the GDL blanket is ongoing.

Objectives

Even though the ﬁeld-scale test has demonstrated that PBs can be used to inject leachate
in MSW landﬁlls, there are no comprehensive design guidelines available for designing
PBs for MSW landﬁlls. Hence, the key objectives of the study presented in this
manuscript are to: (1) prepare preliminary design guidelines for the use of PBs to inject
liquids or recirculate leachate in MSW landﬁlls; and (2) present ﬁeld-scale testing results
of an instrumented PB made up of crushed glass installed at an active MSW landﬁll use

the data to validate the modeling approach presented in this study. PBs can be horizontal

98

or inclined (Khire and Haydar 2003). In this paper, we have focused on horizontal PBS.
In order to accomplish Objective No. 1, we have numerically simulated the effect of these
parameters on the wetted width, rate of travel of injected liquid, and liquid pressure head
in the blanket: (1) hydraulic properties of waste; (2) geometry of PB; (3) hydraulic
properties of PB; (4) settlement of PB; (5) leachate dosing frequency; and (6) initial
degrees of saturation of waste and PB. Wetted width and the rate of travel are key
parameters to design the areal extent of a PB. Liquid pressure head is a key parameter to
evaluate the effect of liquid injection on the slope stability of the landﬁll. Hence, these

parameters were selected to develop the design guidelines.

Advantages of Permeable Blankets

Key advantages of PBS over conventional leachate recirculation methods are: (l)
excavation of waste is not needed during the construction of blanket, resulting in no
odors; (2) A PB can substitute multiple HTS or vertical wells resulting in lower
installation cost from an equivalent design performance; (3) relatively uniform
distribution of injected leachate below the PB resulting in potential reduction in
differential settlement and related post-closure maintenance costs; and (4) PBS made up
of granular materials (e.g., pea gravel, crushed glass, etc.) provide an ideal platform to
embed sensors for monitoring the pressure, temperature and other physical, chemical, or
biological parameters associated with the migration of injected liquids (Haydar and Khire

2005c)

99

LEAHCATE FLOW IN WASTE
Korﬁatis at al. (1984) have demonstrated success when the authors applied Richards’
equation for saturated and unsaturated ﬂow to a lab-scale MSW sample having diameter
and height equal to 0.5 m and 1.5 m, respectively. A wide range (size and composition)
of organic and inorganic constituents in MSW exhibited heterogeneity and anisotropy in
hydraulic properties of MSW. The physical and hydraulic properties of waste may
typically not only vary between landﬁlls, but also within the same cell of a landﬁll. The
saturated hydraulic conductivity of MSW, kw, has been reported to range from 10'8 m/s to
104 m/s (Hughes et al. 1971; Fungaroli and Steiner 1979; Korﬁatis et al. 1984; Oweis et
al. 1990; Bleiker et al. 1993). The ﬁeld capacity of MSW in landﬁlls ranges from 20% to
35% (Oweis et a1. 1990; Korﬁatis 1984; Noble and AmOld 1991). McCreanor (1998)
used SUTRA-ZD to simulate the hydrodynamics of an LRS consisting of a single HT.
The author modeled the lateral and vertical migration of leachate by simulating both,
continuous and intermittent (8 hr on/16 hr oft) applications of leachate. The initial degree
of saturation of MSW was assumed equal to 40%. MSW was simulated as a
homogeneous and isotropic medium using kw = 10's, 10*5 and 10’7 m/s. The modeling
results showed that the lateral and vertical spread of wetted area is almost similar for
continuous or intermittent applications. McCreanor (1998) showed that, the use of daily
and intermediate covers short-circuits the leachate ﬂow along the cover material. The
flow pattern of leachate is also affected by channeling (Zeiss and Ugguccioni 1995).

In this numerical study, we assumed MSW as a homogeneous and isotropic porous
medium having kw ranging from 10'7 m/s to 10'5 m/S. Effect of channeling was not

considered in this study. Even though this assumption may not be completely in line with

100

ﬁeld conditions, the results from this numerical study would be useful in comparing
designs or to investigate alternatives during iterative design phases of an LRS (Straub and
Lynch 1982). Similar approach has been used by Haydar and Khire (2005a) for preparing
preliminary design guidelines for liquid injection using HTS. The effect of heterogeneity
and anisotropy on the hydraulic performance of LRS consisting of HTS is presented in
detail by Haydar and Khire (2005b). Haydar and Khire (2005b) showed that the
introduction of heterogeneity and anisotropy resulted in lower leachate pressure heads for
a given leachate injection rate compared to that when waste was assumed homogenous
and isotropic. This ﬁnding would result in conservative design and can be extended to

LRS consisting of PBS.

NUMERICAL MODELING OF PERMEABLE BLANKETS

HYDRUS-2D Computer Model

Haydar and Khire (2005a) used HYDRUS-2D to simulate the ﬂow of injected leachate
using HTS in bioreactor landﬁlls. HYDRUS-2D is a computer model that can simulate
water, heat, and solute migration in unsaturated, partially saturated, or fully saturated
porous media (Simunek et al. 1999). The program numerically solves the Richards’
Equation for saturated and unsaturated water ﬂow and uses van-Genuchten function (van
Genuchten 1980) for soil-water characteristic curves and van-Genuchten-Mualem model
(Mualem 1976) for predicting the unsaturated hydraulic conductivity function. This
model has been used for saturated/unsaturated liquid and solute transport through porous
media in several studies (Scanlon et a1. 2002; Henry et a1. 2002; Pang et al. 2000; Rassam

and Cook 2002).

101

Conceptual Model

Khire and Haydar (2003) developed a conceptual model to numerically evaluate the use
of PBS as a leachate recirculation alternative to conventional HTS. This conceptual model
is presented in Figure 1. The conceptual model consisted of a PB and a leachate
collection system (LCS). The simulated PB consisted of a 15-mm-thick permeable layer
with a perforated pipe having diameter equal to 0.1 m. The perforated pipe ran
perpendicular to the plane of the paper through the center of the PB. The width of the PB
ranged from 60 m to 150 m (Figure lb). The vertical distance, D, between the PB and the
top of the LCS ranged form 5 m to 20 m. Distance from the top of the PB to the upper
zero ﬂux boundary was assumed equal to 5 m to contain all injected leachate and to
prevent possible artesian conditions for the simulated leaChate injection rates. The LCS
consisted of two 0.15-m-diameter perforated pipes embedded in a 0.3-m-thick gravel
layer at a horizontal spacing, d, equal to 60 m. The slope (tan ﬂ) of the LCS was assumed
equal to 3.5%. The hydraulic conductivity of the LCS drainage material (kLCS) was
assumed equal to 10'2 m/s. The chosen LCS design parameters (tics, krcs, d, and tan B)
resulted in less than 0.3-m-leachate pressure head on the lining system for all simulations
presented in this study.

The simulated saturated hydraulic conductivities of MSW, kw, ranged from 10'7 m/s
to 10'5 m/s. However, an average kW equal to 10'6 m/s was used in most simulations.
These values were selected according to typical values published by Hughes et al. (1971),
Fungaroli and Steiner (1979), Korﬁatis et a1. (1984), Oweis et a1. (1990), and Bleiker et
a1. (1993). The saturated and unsaturated hydraulic properties of the simulated waste, PB

material, and LCS gravel layer input to HYDRUS-2D are presented in Table 1.

102

Zero Flux Boundaries

- . ~. ~ MSW~~"-=  .

 

 

 

  
       
  

 

(a)

 

‘j‘. Leachate
Level

 

lﬁ—10t020m Ap‘

 

i 1 tan ’6 _ I l Liner (Zero Flux
LCS Pipe d ' 60 m Boundaries)

(Seepage Face Boundary)

      
 
 

Permeable Drainage

Perforated LR Materlal

Pipe (Constant Flux. Q)

b 17.17;".11'.” ' -' '-
( ) A 0.45 m
Embedded Sensor Location of reported hp

I ‘ Width ~

l 60 to 150m

(7— Depth~
-- :33 0.15to

 

 

Notes:

hp = Simulated Pressure Head
LCS = Leachate Collection System
LR = Leachate Recirculation

MSW = Municipal Solid Waste

PB = Permeable Blanket

NOT TO SCALE

.O‘SﬁPP’Nr‘

Figure 1. (a) Conceptual model used for numerical simulation of leachate recirculation in
MSW landﬁll using a horizontal permeable blanket; and (b) detail of a permeable

blanket.

103

We simulated leachate as pure water. Hence, any reference to leachate ﬂow henceforth
corresponds to water ﬂow. The results of this study can be applied to any injected liquids
for bioreactor landﬁlls as long as the liquid’s physical and hydraulic properties are
similar to that of water. The effect of gas ﬂow, temperature and biochemical reactions

occurring within a landﬁll was ignored.

Boundary Conditions

All external boundaries were simulated as zero ﬂux boundaries. The perforated pipe used
for leachate injection was simulated as a constant ﬂux boundary. The ﬂux (dimensions:
MOLIT'I) assigned to the boundary was calculated by dividing the leachate injection rate
(dimensions: M°L3T'l) by the perimeter of the pipe for a unit length of the pipe. The
simulated leachate injection rates (Q) ranged from 1.1 to 3.6 m3/hr/m. These rates were
selected based on leachate injection rates used in the ﬁeld for PBS tested in this study and
other studies (Khire and Haydar 2005). The maximum leachate injection rate of 3.6
m3/hr/m corresponded to the maximum rate the pump at the site could deliver for the total
head that existed for the system. Note that Q in m3/hr/m represents the leachate injection
rate in cubic meters per hour per linear meter length of the pipe perpendicular to the plane
of the paper.

Leachate collection pipes embedded in the LCS were simulated as seepage face
boundaries. Leachate ﬂow as a result of percolation from the cap or waste above the
model domain was assumed zero. We believe that this assumption is reasonable because
our key objective in this study was to simulate the subsurface hydraulics of injected

leachate.

104

Table 1. Saturated and unsaturated hydraulic properties used in HYDRUS-2D

 

 

 

 

 

simulations.
Landfill Unit Material 0, 6, a n k,
(l/m) (m/s)
Waste Silt loam 0.078 0.45 3.6 1.54 10'5,10‘and10'7
Pea gravel 0.01 0.3 57.4 2.44 10'2 and 10'3
Permeable
Blanket Crushed 0.02 0.47 12 5 3x102
Glass
Leachate Pea gravel 0.01 0.3 57.4 2.44 10'2
Collection
System

 

 

 

 

 

 

 

105

MODELING RESULTS
Leachate recirculation in the ﬁeld is often carried out in on/off dosing cycles. The dosing
cycles Simulated in this study ranged from 2 hours on/22 hours off to 8 hours on/ 16 hours
off to cover various dosing volumes and frequencies for a typical MSW landﬁll. Leachate
injection was simulated as constant ﬂux under a positive leachate injection pressure.

Figure 2 presents a schematic of the wetting front of injected leachate traveling in
the PB. We deﬁned the wetted width as the distance traveled by the injected leachate
from the injection pipe in the PB. We differentiated between the wetted width of the
underlying waste and the saturated wetted width of the PB as shown in Figure 2. The
wetted width of waste (WW) was deﬁned as the maximum distance traveled by the
injected leachate from the injection pipe in the PB just above the underlying waste. WW
dictates the lateral extent of inﬁltration of injected leachate through the underlying waste.
The saturated wetted width of PB (W3) was deﬁned as the one-half of the width of the PB
where the entire depth of the PB is 100% saturated. W3 is always less than WW. The
difference between WW and W3 varies depending upon the leachate injection rate, the
thickness of the PB and the hydraulic properties of the PB and waste.

When a 60-m-wide blanket is fully saturated, W3 equals 30 m. The injected
leachate also temporarily increases the degree of saturation of the PB and the pressure

head in the PB (hp).

106

~MSW~

Wetting front of
Injected leachate

  

100% Saturation Saturated wetted
R varies)

I width of PB (we, _’

 

 

    

I r\
I
\/ %

 

 

 

 

L Wetted width of waste >
(WW, varies)
Permeable ~ ~ Perforated leachate
blanket (PB) MSW injection pipe

Figure 2. Schematic of wetting front of injected leachate in permeable blanket.

107

Because one of the key parameters that impact the shear strength of the waste is the
effective stress, the pressure head distribution in the blanket is a key input for slope
stability analysis of bioreactor landﬁlls. We have simulated hp by measuring the pressure
head in the proximity of the injection pipe (within 0.5 m) at the bottom of the PB as
shown in Figure lb. The pressure head in the PB is always greater near the injection pipe.
Hence, the use of this pressure head for slope stability analysis would yield conservative
results. Note that hp is different from the injection pressure head in (h,) side the injection
pipe and is a function of the leachate injection rate and the hydraulic properties of PB and
waste. We have presented hp because we also believe that it can be used to interpret and
monitor the hydraulic performance of the LRS. For example, if the magnitude of hp is
close hi, it indicates a good hydraulic continuity between the injection pipe and the PB. If
the difference between h,- and hp increases over time for a given magnitude of leachate
injection rate, it indicates a decrease in the hydraulic conductivity of the blanket or
potential clogging of the leachate injection pipe.

Over 150 simulations using HYDRUS-2D were conducted to evaluate the effect of
the design parameters on the wetted width and pressure head of injected leachate in PB.
The following design parameters were evaluated: (1) hydraulic properties of waste; (2)
geometry of PB; (3) hydraulic properties of PB; (4) settlement of PB; (5) leachate dosing
volume and frequency; and (6) degree of saturation of waste and PB.

Unless speciﬁed otherwise, all simulations were conducted using the following
parameters as input: (1) PB width = 60 m; (2) PB depth = 0.15 m; (3) initial degree of
saturation of PB (SB) = 50%; (4) initial degree of saturation of waste (SW) = 45%; (5)

hydraulic properties of PB are those of pea gravel presented by Khire et al. (2000) with a

108

saturated hydraulic conductivity (k3) = 10'2 m/s; (6) hydraulic properties of MSW are
those of loam from HYDRUS-ZD’S database with a saturated hydraulic conductivity (kg)

= 104’ rn/s; and (7) vertical spacing between PB and LCS (D) = 5 m.

Hydraulic Properties of Waste

Due to the heterogeneity in the composition of MSW, the measurement of representative
hydraulic properties (saturated and unsaturated) of MSW remains challenging till this
day. We evaluated the effect of unsaturated hydraulic properties of MSW on the WW and
hp. In order to do so, MSW was simulated as sand and loam in two separate simulations.
The saturated hydraulic conductivities of sand and loam were assigned equal to 1045 m/s.
However, different soil-water characteristic curves were assigned for sand and loam from
HYDRUS-2D database. In one scenario, leachate was continuously injected at an
injection rate equal to 1.1 m3/hr/m for an injection period of 8 hours and in the second
scenario, leachate was injected in 4 hours on/20 hours off dosing cycles for a total period
of 7 days. The initial degrees of saturation for PB and MSW were equal for all
simulations in these scenarios. The simulation results indicated that unsaturated hydraulic
properties of the waste are critical only if simulations are conducted to evaluate wetted
width and leachate pressure head for relatively short-term (less than 2 to 3 days) leachate
injection scenarios.

For long-term and recurring leachate injection scenarios, which are most typical in
the ﬁeld, unsaturated hydraulic properties of the material used to simulate MSW have
virtually no inﬂuence on the wetted width and the pressure head of injected leachate in
the PB. Hence, unsaturated hydraulic properties of loam were used for all simulations

conducted in this study, unless it is speciﬁed otherwise.

109

Saturated Hydraulic Conductivity of MSW

Figure 3 presents the effect of kw on the simulated W3 and hp in a 60-m-wide PB. The
simulated values of kw were 105, 106, and 10’7 m/s. These values were selected based on
the typical reported values for MSW (Haydar and Khire 2005a). Q was assumed equal to:
(l) 1.1 m3/hr/m for an injection period of 8 hours (Figure 3a); and (2) 3.6 m3/hr/m for an
injection period of 3 hours (Figure 3b).

Figure 3a shows that for kW = 10'5 m/s, the simulated W3 was about 6 in after 8
hours of continuous leachate injection. The simulated hp remained below 0.3 m
throughout the leachate injection period. For kw = 10'7 m/s, W3 reached the maximum
possible value of 30 m for the 60-m-wide PB after 6 hours of continuous leachate
injection. hp rose as the injected leachate traveled within the PB and increased its degree
of saturation. Once the injected leachate reached the 30-m-distance and it saturated the
entire blanket, hp sharply increased as the storage capacity of the blanket was exceeded. It
was also noted that this increase in the pressure head was more for a lower hydraulic
conductivity of the underlying MSW.

Figure 3b Shows that for kw = 10'5 m/s, the simulated W3 was about 17 m after an
injection period of 3 hours for Q = 3.6 m3/hr/m compared to about 6 m for Q = 1.1
m3/hr/m after an injection period of 8 hours. The simulated hp was about 1.5 111, whereas
it was 0.3 m for the lower Q. For kw = 1045 and 10'7 m/s, W3 was 30 m after an injection
period of 2.5 and 1.5 hours, respectively. For Q = 3.6 m3/hr/m, the hp values were greater

than those for Q = 1.1 m3/hr/m.

110

      
  
    
 
   
  
   

 

7 WIIY'YYII‘;.YTIITTTIITITTIIIIIYV I 0.1
l .
q

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

'3 25i— f/i: =10‘5m/s ..1 "5g
3: i : “f . ‘ 'U :Q
a? l 5 la 5 l to -
3 .. ‘— ikw=10 m/s : . o m
3 g l 5 .E
N - ~ 4
5 m l— : ' m 3
33°- 15 if k =10'7mIs ‘ 82
in '5 L W ‘ n. 0
13 5 “r ' *L : ﬁ-i '81.:
*3; 10 '5 . Wettedwidth? —‘i 10 :5 .8
- :L _ . . 1 I 3
E r kw-106m/s g -—Pressure Head J E g
s» , - "M10 "V5 i o=1.1m3/hr/m,
i / B=1O'2m/S.Sa=50% and sw=45%
0 - «wee - 100
0 1 2 3 4 5 6 7 8
Time since Leachate Injection Started (Hours)
35 r . r . 1 . I . 0-1
1 - .
| Wetted Width Pressure Head 03) j
30 * : ; ' . E
.8 l r z / -e / d “6 VG.
% A 25 k =10'7m/s . k ‘10 m/s r g c
a E i /z a 1 35 g
3 gm - 3 i‘k '1O5m/s -—.i O C
H 20 ‘— -/ 5 .-
E " i : ﬂ 1» ID
3 E i; 5 3wk =1043 rn/s _ﬁj 3 *5
(U u- l” h .C
to o 15 p n. g
I: .—
36 i iwg3
.12 § 10 :— . . j g 1:
:1 t r g 5 _ ,7 a 3
_g l kw =10'5m/s : gkw'm "V5 —-l g 8
W 5 f : I . ' ‘0 'E
l /Q= 3.6m/hr/m. k8 =10'2 m/s SB =50%and sw=45% j
0 . . . e - 100
0 0.5 1 1.5 2 2.5 3

Time since Leachate Injection Started (Hours)

Figure 3. Effect of kw on simulated wetted width and pressure head of injected leachate
for a 60-m-wide PB for Q— = 1.1 m3/hr/m (a); and Q= 3. 6 m3/hr/m (b).

111

Unsaturated Hydraulic Properties of Waste

Measurement of the unsaturated hydraulic properties of MSW remains challenging till
today because these properties can vary signiﬁcantly within a landﬁll or a landﬁll cell.
However, these properties influence the estimation of WW and 11,. To simulate the effect
of unsaturated hydraulic properties on WW and hp, soil water characteristic curves
(SWCCS) of loam and sand were used to simulate waste, while maintaining the saturated
hydraulic conductivity of the waste constant to allow for comparison. Figure 4 presents
the effect of waste SWCC on WW and hp in a 60-m-wide PB for kw = 104’ m/s. Q was
assumed equal to 1.1 m3/hr/m for a dosing frequency of 4 hours on/20 hours off. Figure
4 shows that W W and hp were greater when waste was simulated as sand compared to
loam. This was because sand has a lower unsaturated hydraulic conductivity than loam.
For the remainder of this study, SWCC of loam has been used to simulate waste.

Designers should consider a SWCC close to that of the targeted waste in the design.

Saturated Hydraulic Conductivity of Permeable Blanket

The saturated hydraulic conductivity of PB (k3) may vary depending on the permeable
material used for constructing the PB. The effect of k3 on W3 and hp was evaluated using
k3 values equal to 110'2 and 10'3 m/s. Figure 5 presents the effect of k3 on the simulated
W3 and hp in a 60-m-widePB for kw = 1045 m/s. Q was assumed equal to l.l m3/hr/m for
an injection period of 8 hours (Figure 5a) and 3.6 m3/hr/m for an injection period of 3

hours (Figure 5b). Figure 5 shows that the injected leachate travels at a slower rate for a

112

 

I i. x I T ' T l I II 1
I a .'

' Waste = Sand
(Solid)

1 .1 i ' 1 Air r ‘ 1 ‘.. T
a ,' 'l
n l
' Waste = Loam
(Dashed)

       
  
   

P

 

 

 

     

 

Waste = Sand

 

 

Wetted. Width

 

 

 

  
  

 

 

 

Pressure Head

 

Dosing Cycle (4 hours on/ 20; hours off),

, - Q = 1.1 m3/hr/m,

_ -6 _ -2 _ _

2,.kw-10 m/s,kB-1O m/s,SB-50%&Sw-45% :

101 ...... -.......4 ........ ALJQL,.11O
0 1 2 3 4 5 6 7 8

Time since Leachate Dosing Started (Days)

Simulated Pressure Head

Simulated Wetted Width of MSW, WW (m)
N
O
l
of Injected Leachate in PB, h (m)

 

 

Figure 4. Simulated wetted width of waste and simulated pressure head in the PB as a
function of the soil waster characteristic curve used in simulating the waste.

113

 

l.-ﬁ—YL.111 TT—T—Yl rrrrrrrrr 0.1

 

 

 

 

 

 

 

 

 

 

 

 

     

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
 
 
  

 

 

———— Wetted Width ; (a) A
. ; “ E
'8 Pressure Head ‘ Va
3:: ‘ ‘ 4 1: '5
E A ~ ° 0
g v ‘ 3 . A I n.
'5 gm L i I # 1 2 C
a, ..
E _ f k =10'2 m/s 3 o
in _ B ‘0 1'6
*3: 2 15 .__ / i/ E f.
to o . / a
g .5 ; k =1o" m/s 1 '3 3
7'3 :2 10 L . B , ' 10 E U
'3 3 i ' 3 §
'- L ' I 3 a '-
‘D 5 i Q =1.1 m3/hr/m, . é : f
l kw = 10“ m/s, $13 = 50%, and sW = 45% . °
0 ....L. .....- - -. . - -1 100
0 1 2 3 4 5 6 7 8
Time since Leachate Injection Started (Hours)
30 T.-L__L ,— .——r--—-i——-'-—~g—~~i 0-1
—— Wetted Width (b) ...
- i E
u 4 V
g 25 j Pressure Head .8 £0.
41 A " . i ‘3’ '
E : m
3...“ k =10‘2m/s 41 f,“-
3 g 5 ~ 5 5
a _ : Q
g3 ~ . . i/ 72’ ‘6
‘6... *k=10'3m/s i i/i 1 if:
m ° 15 i B " "k -10'3m/s : '° 8
'u 5 . _ ‘ 1 o ..l
3 'U . / B ; 7 "j 10 a 'U
2 g _ . f 3 2
:3 . j . , . E 0
g 10 P ‘ ’ i 1 5 ‘2
(D Q = 3.6 m3/hrlm, ; ‘ . é
: kw = 1045 m/s, 83 = 50%. and Sw = 45% : o
5 _____4__ A g 1 l 1 1 1 1 * 100

 

 

 

o .05 1 1.5 2 2.5 3
Time since Leachate Injection Started (Hours)

Figure 5. Effect of k3 on simulated wetted width and pressure head of injected leachate
for a 60-m-wide PB for: (a) Q = 1.1 m3/hr/m; and (b) Q = 3.6 m3/hr/m.

114

lower kg and results in a greater hp for a given leachate injection rate. Thus, a greater k3 is

preferable when selecting a suitable material for the PB under consideration.

Hysteresis in Hydraulic Conductivity of Permeable Blanket

The effect of hysteresis in kg on WW and h,, was evaluated by simulating the SWCC of the
PB for wetting and drying, one at a time. HYDRUS-2D model incorporates hysteresis in
the soil hydraulic properties. When a hysteretic description of the soil hydraulic
pr0perties is selected, a selection must be made for whether the initial condition is
associated with the wetting or drying SWCC. For this evaluation, the simulated PB
material was gravel (Table 1). The hydraulic parameters of the drying SWCC are
presented in Table 1. The HYDRUS-2D model generated the wetting curve of the SWCC
by multiplying the a parameter in the drying curve by a factor of two. This resulted in
assigning on = 114.8 for the wetting SWCC. For this evaluation, a 60-m-wide PB was
simulated. kW was assumed equal to 1045 m/s and Q was assumed equal to 1.1 m3/hr/m for
an injection period of 8 hours when liquid was injected continuously. The simulated
values of SB and S W were equal to 50% and 45%. Simulations showed that hysteresis in kg

had no inﬂuence on WW and hp.

Geometry of Permeable Blanket

Permeable Blanket Depth

The depth of a PB can be inﬂuenced by the material used for constructing the PB. For
example, a GDL blanket can be relatively thin (~ 10 mm), whereas blankets made up of
shredded tires can be as thick as 0.6 m depending on the size of the tire shreds. Figure 6

presents the effect of PB depth on W3 and hp in a 60-m-wide PB. The simulated PB depth

115

 

Tli‘lj' 7T llrjl—Ti fit? flﬁrl IIITTITITT 0.1

    
 

 

 

 

 

 

 

 

  
 

 

 

 

 

   
 
 

:0: 1.1 m3/hr/m, k =10’5 m/s, k :102 m/s A
W B E
'3 25 1' . . _ 0.45-m-deep PB 5 q
3: _ : I 3 ; : : 4 'U C
E E - 5 5 . ~ I 3 \‘ . 8 m-
'o m 20 3“ i 3 5 -/ s 3 ? - 1 a, 0-
o g ;- s 2 0.15-m-deep PB 5 ; : .. ,5
a " 7- | _‘h i . 1 . ' d a a
5 E 15 _ —-——Wetted Width ' 0.15m-deep PB . g E
(75 “6 : Pressure Head I l g 8
'o .l: - - - 1 °
- 3 I i = 3
a .
E 0.45-m-deep‘ PB g g
27: 5 ; 2 ; z 1 ”’ E
, 1 ‘ 88 = 50%, and SW = 45% 3 “6
O :4 L.._L.._- 1-4... ..;__.L.___..L_.'_J....L.. l.._l..l __L__L_i ' L 1 l l l 1 J_L 1 l 1 1 1 L 1 J J_L...1,_:j 100

 

 

0 t 2 3 4 5 6 7 8
Time since Leachate Injection Started (Hours)

Figure 6. Effect of PB depth on simulated wetted width and pressure head of injected
leachate for a 60-m-wide PB.

116

values were assumed equal to 0.15 m and 0.45 m. kw was assumed equal to 10'6 m/s and
Q = 1.1 m3/hr/m for an injection period of 8 hours. Simulation results presented in Figure
6 indicate that W3 decreases as the PB depth increases. However, WW did not differ
signiﬁcantly between the two simulated depths of the PB (W W not shown in Figure 6). As
the depth of PB increased, the storage capacity of the blanket also increased. This
increase in storage capacity resulted in a lower h,, for the deeper PB. These results are
consistent with a ﬂow problem where two circular pipes having different diameters are
connected in series. The liquid pressure is less in the larger diameter pipe. Thinner
blanket is preferable to keep the capital cost down, whereas thicker blanket is preferable

to keep hp in an acceptable range for slope stability concerns.

Permeable Blanket Width and Vertical Spacing

The planar dimensions of PBs may vary according to the landﬁll cell shape and its
leachate recirculation needs. PBs can be installed at any ﬁlling stage of a landﬁll before it
is capped. Thus, PBs can be installed at various surface elevations of an MSW landﬁll.
We simulated 60-m-wide and ISO-m wide PBs for kw equal to 10'6 m/s and Q = 1.1
m3/hr/m for an injection period of 8 hours. The simulation results indicated that the PB
width had no effect on the simulated wetted width or h,,. We also evaluated the effect of
vertical spacing between the PB and the LCS by simulating a vertical spacing (D) equal
to 5 m and 20 m. kw was assumed equal to 1045 m/s and Q = 1.1 m3/hr/m for an injection
period of 8 hours. The simulation results indicated that D did not influence the simulated

wetted width or hp,

117

Settlement of Permeable Blanket

Settlement of waste in landﬁlls has been reported to range from 10% to 30% of its initial
thickness (Sowers 1972; Wall and Zeiss 1995). The composition of waste, climate,
presence or absence of leachate recirculation, and other physical and biochemical factors
impact the settlement of waste in landﬁlls. Due to the differential settlement of waste,
PBs constructed within waste will also undergo differential settlement.

The effect of settlement of waste on the leachate ﬂow through a 60-m-wide PB was
evaluated by simulating 3-m-deep sag at the center of the PB. The 3-m-deep sag
simulated differential settlement equal to 3 m. The simulation results indicated that for
leachate injection pressure heads less than the magnitude of the simulated sag (3 m in this
case), the injected leachate could not ﬁll up the entire blanket. Hence, in such cases,
corresponding leachate injection rate needs to be increased to create an additional
injection pressure head that can compensate for the sag. Once the injection pressure head
was greater than or equal to the magnitude of the sag, the injected leachate ﬁlled up the
entire blanket resulting in a greater WW and W3. Haydar and Khire (20050) have
presented a sensing and monitoring system that can be used to monitor the leachate
injection rate and pressure and the rate of travel of injected leachate in the PB. Such
sensing and monitoring system can be used to manipulate the leachate injection rate (or

pressure) to compensate any differential settlement in the PB.

Leachate Dosing Frequency

In order to prevent buildup of liquid pressure head and leachate breakouts, and possible
slope instabilities, leachate is not continuously injected in landﬁlls. Instead, leachate is

injected in on/off dosing cycles. The dosing volume and frequency for leachate injection

118

may vary depending on the daily leachate generation volume and the operational needs of
the landﬁll. We evaluated the effect of dosing frequency for Q = 1.1 m3/hr/m on WW by
simulating leachate injection for 2 hours on/22 hours off, 4 hours on/20 hours off, and 8
hours on/ 16 hours off. Figure 7a presents the effect of dosing frequency on W for kw =
1045 m/s and SW = S3 = 30%. Figure 7a indicates that WW is a function of the ratio of on to
off leachate injection duration. The wetted width of waste was greater for a dosing cycle
where the on to off times ratio was greater. For a given dosing frequency, WW increased
as the number of leachate dosing days increased until WW reached a constant maximum
value ranging from 15 to 45 m depending on the on to off duration ratio after almost ten
days (Figure 7a).

Figure 7b presents the simulated hp for the dosing frequencies of 2 hours on/22
hours off and 4 hours on/20 hours off for the same simulations presented in Figure 7a.
However, for a given dosing frequency, hp increased and then remained almost constant
as the leachate dosing continued and the system almost reached a steady-state. The initial
increase was due to the increase in the degree of saturation of the waste and the PB. The
magnitude of hp was a function of the on to off duration ratio. Greater the on to off
duration ratio, greater was the magnitude of hp.

Even though it is not customary to continuously inject leachate in landﬁlls, we
evaluated the effect of continuous leachate injection (until steady-state is reached) on WW
to obtain the maximum possible wetted width for a given leachate injection rate. Figure 8
shows a contour map that presents simulated WW for a 150-m-wide PB using continuous

leachate injection at Q = 1.1 m3/hr/m (Figure 8a) and Q = 3.6 m3/hr/m as a ﬁmction of kW

119

A ._, .7 .L. L - __ _ 4 _ .iwL IL”- 2.? 2,22 _L_L,___ ‘1
g; . (a) 1
s g
a 40 l— 1
E 8 Hours On/16 Hours Off j
“5 1
5 3° ” ‘1’
:2
J
E 4 Hours On/20 Hours Off .3
a 20 ~ ‘47
3.1 i
«1 ’ﬁ.
3 2 Hours On/22 Hours Off _1
g g 0 =1.1 m3/hr/m, kw=1045 m/s, kB :102 m/s,1
.g ; 88 = 30% and sW = 30% i
(D 0 _ __ ._. ._l - .= 1 1 1 1 1 ___;__#d__._.l
0 2 4 6 8 10 12 14
Time since Leaching Dosing Started (Days)
6 1 [ _ T T 'T‘ "T" “Tm" "1 """ T T" "‘"T' W "‘T’“ "YT—T" :
3 (b) :
i i l I l I l l 1
.E A 1* 4 ‘11 mm H 011(5 I'd L' ) ‘
- E . ours ours on me .
“5.1 l— , n u u u n n n n :
'3 c 05 1. l 2 Hours On122 Hours Off (Dashed Line) -7
g '~ I : 2 ‘1 - . 1 . i ' i j
I m I , l . , , l ' ' I . . ' :
o a. ll ' 1 i ‘ . ' 3 1 3 2 I 3
5 .E - : J 1 I ' i ' ' .‘ ‘ 1 -
g 2 3 1 3 1 1 1 1 1 1 1 i 1 :
9% iiiiiiiiiiiiiii
l I
$3 0. '1 "i ‘K\‘,‘-."-.‘-.'~"-‘
0 g. I I. - I I‘
la" Xi'.‘ ‘11": '1 ‘0 5 1.1.1.13
g 1, '- ' Q = 1.1 malhr/m, kW= 10"5 m/s, k8 = 10'2 m/s,
“0'1 TE 3 = 30% and s = 30%
l B W
.0.5 E l L l 1 l J 4 4 l J
0 2 4 6 8 10 12 14

50

 

 

 

 

 

 

 

 

 

Time since Leaching Dosing Started (Days)

Figure 7. Simulated wetted width of waste as a function of leachate dosing frequency for
a lSO-m-wide PB (a); and corresponding simulated pressure head in the PB (b).

120

 

_‘5‘___
_«__r

 

 

 

 

_1_ L
ﬂ
m

 

 

 

 

 

 

”N4“ L.
”...—‘4

 

 

 

k B (CM/S)
10 ‘

 

I ‘ Wetted wldth :5; ll /
q ofwaste(m) ’ / 8833$8

 

 

 

 

\

Wetted wldth :9
i of waste (m) 7:!

k3 (cm/s)
10'1

 

 

L—K
.4
L.\\
r Y

 

Figure 8. Contours of Simulated maximum wetted width of waste at steady-state as a
function of kg and kW for continuous leachate injection rates: (a) Q = 1.1 m3/hr/m; and (b)
Q = 3.6 m3/hr/m.

121

and k3 (Figure 8b). kw was varied from 10'7 m/s to 10'5 m/s and kg was varied from 104
m/s to 10'2 m/s. The simulated WW was mainly a function of kw as long as [(3 was greater
than kw. The simulated maximum WW increased as kw decreased. The key ﬁnding ﬁ'om
Figure 8 is that either increasing Q or reducing kw can increase WW. Because kw in the
ﬁeld typically cannot be controlled, Q is the key parameter that needs to be controlled to

achieve a certain WW.

Degrees of Saturation of Waste and Permeable Blanket

The rate of travel of injected leachate in a PB is a function of the initial degrees of
saturation of waste (SW) and the PB (S3). The initial degrees of saturation are input to
HYDRUS-2D as an initial condition in the form of watercontent or matric suction head.
Figure 9 presents the effect of initial conditions (SW and SB) on the simulated WW for a
lSO-m-wide PB for kw = 1045 m/s and Q = 1.1 m3/hr/m for a dosing frequency of 4 hours
on/20 hours off. The x-axis of the plot in Figure 9 represents the number of days since
leachate dosing started. The y-axis represents the maximum WW at the end of the daily
leachate injection event. We assumed four possible sets of initial conditions with SW
ranging from 30% to 65% and SB ranging from 30% to 95%. This range of initial degrees
of saturation was selected based on the typical range of values observed in the ﬁeld for
bioreactor landﬁlls (Zhao et al. 2004; Haydar and Khire 2005c). Figure 9 shows that for
all four sets of simulations, WW increased as the number of leachate dosing days
increased. However, irrespective of the initial degrees of saturation of the waste and the
PB, the maximum WW was about the same after a few days (long-term) of leachate dosing
for a given value of Q and dosing frequency. It took progressively longer time to reach

the maximum W w for lower values of S w and SB (Figure 9).

122

 

 

 

 

 

g 30 — e. e _ ...-.- -__.___._,
3 f i . V o . <> : ’
g I ‘2’2‘; A A :424243
.‘ * A 9 I ‘ - 3’; 3; - i
. 3 24 r ., z - . /="" = _.
m it . é : , "3’ l
2 2» Z ' 1
“o- l ’/ / ' ; 1'
5 18 6 - c —9— Se = 30% and SW = 30% l
E 5 . —a—sa=95% and sw=30% l
E P —-e— 88 = 30% and sw = 65% {
0 12 ° ' , —a—s =95%andS =65% J:
o . ,_
3 I» 1
' J
3 .L i ? : 1
‘15 6 + Q = 1.1 m3/hr/m (4 Hours On/20 Hours Off), j
3 k =1o’6 m/sandk =10'2m/s
.g l: W B j
(I) 0 L___.___ _____*i______r__;__ ._ﬁwrﬁﬁ . i i 1

 

_s
.b

0 2 4 6 8 10 12
Time since Leaching Dosing Started (Days)

Figure 9. Simulated wetted width of waste as a function of the initial degrees of
saturation of the waste and permeable blanket for a lSO-m-wide PB.

123

Initial Degree of Saturation of Waste

The degree of saturation of waste plays a key role in operating an efﬁcient bioreactor
landﬁll. The effect of initial degree of saturation of waste was evaluated by simulating S w
values equal to 30%, 45%, and 65%. Porosity of the simulated waste assumed equal to
0.45, the simulated Sw values corresponded to volumetric water contents ranging from
0.15 to 0.3. Figure 10 presents the effect of SW on the simulated W3 and hp in a 60-m-
wide PB for kw = 10*5 m/s and Q = 1.1 m3/hr/m over a period of 8 hours of continuous
leachate injection. SB was maintained at 50% for all simulations. Figure 10 shows that
greater the SW, faster the rate of travel of the injected leachate. When the simulated
leachate injection was continued beyond 8 hours (not shown in Figure 10), W3 and WW
reached a maximum value that was about the same for various S... values indicating that a

steady-state was reached.

Initial Degree of Saturation of Permeable Blanket

The degree of saturation of the PB varies and depends upon the rate, duration and
frequency of leachate dosing cycle, inﬁltration of precipitation, and the hydraulic
properties of the surrounding waste and PB. Haydar and Khire (2005c) reported that SB
varies signiﬁcantly over the operational life of the PB. Figure 11 presents the effect of SB
on the simulated W3 and hp in a 60-m-wide PB for kw = 10'6 m/s and Q = 1.1 m3/hr/m
during an injection period of 8 hours. The simulated values of SB were equal to 30%,
65%, and 95%. Note that the initial degree of saturation of waste (SW) was maintained
constant at 45%. Figure 11 shows that the injected leachate traveled faster for a greater

S3. However, the difference in the rate of travel of injected leachate for the various

124

     
  
 
 

 

 

l; 11 irrlrvrf70.1
—-—WettedWidth

 

 

 

 

 

'1
— Pressure Head 4;

 

 

 

 

Simulated Saturated Wetted
Width of PB, W (m)

30 =1.1 malhr/m, kw =1o" m/s,

   

Simulated Pressure Head
of Injected Leachate in PB, hp (m)

QkB = 10'2 m/s and S8 = 50%

o JJIAAJLJLLLllliLLiJ‘111—LlilllJllllliiL4l‘100
O 1 2 3 4 5 6 7 8

Time since Leachate Injection Started (Hours)

 

LLA .l.....L. ... - .. .._.__.. . , ..l 1-..... 1 .. 1 ....

 

Figure 10. Simulated wetted width and pressure head of injected leachate as a function of
the degree of saturation of waste (SW) for a 60-m-wide PB.

125

 

30 1:111 r Trrﬁ'ﬁr YﬁjﬁTT‘TIY -T;TTIYTYTY'j 0.1
~ . = o _T W“

:33 330/" : -———- Wetted Width j

/:&SB=50% z ' '

 

 

 

 

 

 

— Pressure Head

p

 

       

10}

Simulated Pressure Head

Simulated Saturated Wetted
Width of PB, W8 (m)
z;
of Injected Leachate in PB, h (m)

 

 

0:;.-._ ................................. j100
0 1 2 3 4 5 6 7 8

Time since Leachate Injection Started (Hours)

 

Figure 11. Simulated wetted width and pressure head of injected leachate versus leachate
injection period as a function of the initial degree of saturation of blanket (S3) for a 60—m-
wide PB.

126

values S3 was relatively small. Similarly, the simulated hp did not differ much for the

various values of SB.

FIELD-SCALE TESTING OF PERMEABLE BLANKET

Field Installation of PB

In order to validate the use of PBs, a horizontal instrumented PB was constructed at an
MSW landﬁll located in Jackson, Michigan. The landﬁll is currently active and generates
on average 45 m3 of leachate per day. The blanket installed is 60—m-long, 9-m—wide, and
0.15-m-deep. The PB is made up of crushed recycled glass having average particle
diameter, D50 = 10 mm and hydraulic conductivity equal to 3 X 10'2 m/s. The blanket was
constructed using the following sequenced procedure: (1) waste surface was leveled; (2) a
non-woven GT fabric was laid on the waste surface; (3) about 0.15-m-deep layer of
crushed glass was placed on the GT; (4) an additional GT fabric was placed above the
glass; and (5) about 3-m-deep waste was placed on the upper GT fabric before leachate
recirculation was turned on. Below the blanket, from top to bottom, there is: (1) about 50-
mm-deep silty soil (loess) used as a daily soil cover; (2) about 20-m-deep MSW; and (3)
leachate collection and lining systems.

A 9-m-long perforated high density polyethylene (HDPE) leachate injection pipe
having an internal diameter (ID) equal to 75 mm was installed at the center of the
blanket, parallel to the width of the PB. One end of the perforated pipe was capped and
the other end was connected to a hydraulic pump followed by three interconnected
leachate storage tanks having a total storage capacity equal to approximately 115 m3. A

leachate ﬂow control valve, a digital pressure gauge, and a magnetic ﬂow gauge were

127

installed in the leachate pipe to control and monitor the pressure head and ﬂow rate of

injected leachate.

Automated Monitoring System

An automated sensing system was installed to monitor the hydraulic performance of the
PB. The glass blanket was instrumented with the following sensors: impedance moisture
content sensors; time domain reﬂectometry (TDR) moisture content sensors; vibrating
wire pressure transducers (piezometers); and thermocouple and thermistor temperature
sensors. These sensors were used to monitor the travel of injected leachate in the blanket.
No sensors were installed in the underlying waste. A detailed discussion on the
monitoring system is presented in Haydar and Khire (2005c).

All sensors including the leachate ﬂow gauge and pressure gauge were connected
to a data logger located at the site. The data logger was programmed to take readings at 1
to S-minutes frequency to allow relatively precise monitoring of the injected leachate and

h,,.

Field Data versus Simulated Results

During the period from September 2003 to April 2004, about 3,200 m3 of leachate was
injected in the glass blanket corresponding to approximately 90 leachate recirculation
events at leachate injection rates ranging from 1.1 to 3.6 m3/hour per meter length of the
embedded injection pipe. The maximum leachate injection rate equal to 3.6 m3/hr/m
corresponds to the maximum rate the pump at the site could deliver for the total head that

exists for the system.

128

Because hydraulic conductivity and degree of saturation of the waste underlying
the blanket are unknowns, in order to simulate the wetted width measured in the ﬁeld, we
simulated WW for the 60-m-wide glass PB for kw ranging form 10'7 m/s to 10'5 m/s and
for SW = 45% and 65% to cover the most commonly reported range of kW. We simulated
Q = 1.1 m3/hr/m because a majority of leachate injection events corresponded to this rate
at the site. The saturated and unsaturated hydraulic properties of the PB measured for
crushed glass in the lab were input to HYDRUS-2D (Table 1). For all simulations, S3 was
assumed equal to 50% based on the measured water content of the PB. Separate sets of
simulations were conducted for SW = 45% (Figure 12a) and 65% (Figure 12b). Figure 12
presents a strong correlation between the WW measured by the sensing system in the ﬁeld
for leachate injection events conducted using Q = 1.1 m3/hr/m in Sep. and Oct. 2003 and
WW simulated for the measured and assumed sets of input parameters. Due to leachate
recirculation during Sep. 2003, the degree of saturation of the waste increased during
Sep. to Oct. 2003 period. Hence, as per Figures 9 and 11(b), the injected leachate traveled
at a faster rate during the Oct.

hp values measured in the ﬁeld in the close proximity of the leachate injection pipe
[Figure l(b)] were simulated for the blanket for three leachate injection events
corresponding to Q = 1.1 m3/hr/m (Figure 13). The rate of increase of hp during a leachate
injection event is primarily a function of SB, SW, kg, and kW. Soil water characteristic
curves (SWCCS) for the underlying waste and PB and hysteresis in the SWCCS play a
relatively minor role for accurately simulating hp during a leachate injection event,

In the ﬁeld, we believe that SW most likely increased during the period from 24 Sept.

2003 to 27 Feb. 2004 due to continuing leachate recirculation. Similar to observed in the

129

30 _ 72., D [-2- .aDaﬁm-.. _-. ,-- ._ -.., fad a

 

 

 

 

 

 

 

 

. Hydrus-ZD: 0 =1. 1 m 3,/hr/m (a) j

.5, 25 f.» 88- - 50% and sw- - 45% j
k 7 -
g . 0 Field Data - Sep 2003 / j
g. 20 Cl Field Data - Oct 2003 kw = 104 m/s -1
U) T 1
2 /
"6 15 ~ 1
5 I
2 ;- 1
3 10 L J
‘0 9 1
O - ..4‘
‘5 1
3 5 z: i
f l
OLA-ea—J—t—L—a—egweuutm ..DLB..21

0 0.5 1 1.5 2 2. 5 3

Time since Leachate Injection Started (Hours)

30 [V "" Y 7“”; "“7 “rﬁr T 'rﬂ‘r v- .V {._fY.. ﬂu”?

,, Hydrus-2D o- -1. 1 m3/hr/m (b);

25 1: SB- - 500/0 and SW- '-' 65°/o _,

0 Field Data - Sep 2003 kw = 10'7 m/s 1

20 :1 Field Data -oct 2003 /

O ;

Wetted Width of MSW, wW (m)

 

 

k = 10’<5 m/s
15 F W a
10 5 .3
kW= 10' m/s "
5 ~ -
0 L...l_._l__...._L_ l ._1_-L_ _1 _i,,_a_ ..._J-_1. .l_a__.i_._a__1_._§t__a ..L. 1.. .__1_‘_r_1_-4__ .1
o o. 5 1 1.5 2 2.5 3

Time since Leachate Injection Started (Hours)

Figure 12. Correlation between the wetted width of waste measured in the ﬁeld and the
simulated wetted width of waste for a 60-m-wide PB for kw ranging from 10'7 m/s to 10 5
m/s, S3 30%, Q=1.1m3/hr/m, SW: 45% (a), and SW: 65% (b).

130

 

10.. T ...... ..ﬂ...e.f
i Q=1.1m3/hr/mandk8= 3x10'2 m/s

.4. ._4....J . .L_L_J..J

’1

 

'0 l

8 g i W- B - O l

0-s O. ,— _3

5 c 1

“a - 1

DE : I. IIIIIIIIIIIIIIIIIIIII

a i .- -..-— - 'VryvV'VVVVWr

E .051 ! aqn‘d'ﬁ'iﬁnﬁnwﬁ'“? D

2 ‘5 fl: 5 E] D D C] a 1:1 D D D D 1:1 :1 U_D_El.IZI_D_ElEl_E_1_D_El Pup]

3 .5 0'1 ' a .-"'.."-: .... -5 i

a 3 D .0 ' Simulated-kW=1O m/s

m I

a ," _Simulated-kw=10'6m/s
.‘S = 65% ands = 50% Field Data 24 Sep 2003 j
.' w 3 Field Data -1 Oct 2003 1

0.01 ' 242. 1 2 1- . . F , Fgelg Data-1275:1113 2094 ll
0 0.5 1 1.5 2 25

Time since Leachate Injection Started (Hours)

Figure 13. Measured and simulated pressure head of injected leachate in the glass
permeable blanket for three leachate injection events corresponding to Q— = 1.1 m 3/hr/.m

13l

ﬁeld, simulated data (Figure 13) indicated that as SW increased, hp also increased. The
magnitude of simulated maximum hp was less for a greater kw. The maximum hp in the
ﬁeld measured by the pressure transducer remained below 1.5 m, which is in the same
range of hp simulated by assuming kw equal to 10*5 and 10'5 m/s presented in Figure 3.

Because kW, SW, and SWCCS of the underlying waste are unknowns, the ﬁeld data
collected in this study cannot be simulated any more accurately. However, Figures 11 and
12 indicate that the data measured in the ﬁeld is in the ball parks of simulated results for
the assumed input. Hence, we believe that the modeling approach presented in this study

is appropriate to design PBs.

SUMMARY AND PRACTICAL IMPLICATIONS
This paper presents a numerical evaluation of key design parameters that are used for the
design of leachate recirculation system consisting of PBs. The saturated-unsaturated ﬂow
model HYDRUS-2D was used to simulate the effect of these parameters on the hydraulic
performance of PBs: (1) hydraulic properties of waste and PB; (2) geometry of PB; (3)
settlement of PB; (4) leachate dosing frequency; and (5) degrees of saturation of waste
and PB. In addition to the numerical modeling, data from an instrumented ﬁeld-scale PB
made up of crushed glass was used to evaluate the application of PB for leachate
recirculation. The key ﬁndings of this study are as follows.
0 In order to maintain the liquid pressure build up in the blanket as little as possible and
to achieve the greatest wetted width, it is important to select a material having the
highest possible hydraulic conductivity to construct a PB. Thus, gravel is preferable

to sand.

132

Increase in the PB depth decreases hp. Hence, a thicker PB is preferable if slope
stability evaluation of the landﬁll requires lower pressure heads in the blanket. A
thicker blanket, however, does not result in a greater wetted width of waste and hence
does not offer greater wetting of the underlying waste.

Greater the hydraulic conductivity of the underlying waste, lower the wetted width
and lower the hp.

The coarser the SWCC of the waste, the greater WW and 11,,

The greater the degrees of saturation of the waste and/or the PB, the faster the rate of
travel of injected leachate in the PB and greater the hp. When leachate is injected in
on/off dosing cycles, SW and SB increase until a steady-state approaches. WW and 11,,
are directly proportional to the on to off duration ratio and the magnitude of the liquid
ﬂux during the on period. Hence, on to off duration ratio needs to be selected based
on site-speciﬁc factors to balance the beneﬁt of a greater wetting volume for a greater
on to off duration ratio to a lower factor of safety against slope stability due to an
increased hp.

If the blanket settles, a greater leachate injection rate (or head) is needed to
compensate for the loss in elevation head to maintain the same WW.

Even though not speciﬁcally evaluated in this study, we recommend designers to

consider the following key design-related issues: (I) maintain sufﬁcient distance (> 15 m)

between the edges of the blanket and the side slopes of the landﬁll to minimize the

potential for leachate breakouts (Haydar and Khire 2005b); (2) inject leachate in

quantities or rates that will not jeopardize the slope stability of the landﬁll due to an

increase in the liquid pressure head; (3) instrument the blanket and monitor the long-term

133

performance of the blanket for potential clogging, settlement, or physical integrity
jeopardizing the hydraulic performance of the blanket; and (4) more closely monitor the
liquid pressure head at critical locations in the LCS to make sure that the LCS is

appropriately draining the injected liquids.

ACKNOWLEDGEMENTS

The project has been jointly funded by Environmental Research & Education Foundation
(EREF) and Waste Management Inc. We sincerely appreciate the support from the
sponsors. The authors are grateful to Dr. Jirka Simunek for his input related to running
the computer model HYDRUS-2D. However, the results and opinions presented in this

manuscript are those of the authors.

134

REFERENCES

Bleiker, D.E., McBean, E., and Farquhar, G. (1993). “Refuse sampling and permeability
testing at the Brock West and Keele Valley Landﬁlls.” Proceedings of the Sixteenth
International Madison Waste Conference, 548-567.

Fungaroli, A. and Steiner, R. (1979). “Investigation of sanitary landﬁll behavior.” Vol. 1,
Final Report, US. EPA 600/2-79-053a.

Henry, E.J., Smith, J .E., and Warrick, AW. (2002). “Two-dimensional modeling of ﬂow
and transport in the vadose zone with surfactant-induced ﬂow.” Water Resources
Research, Vol. 38, No. 11, 331-3316.

Hughes, G., Landon, R., and Farrolden, R. (1971). “Hydrogeology of Waste Disposal
Sites in Northeastern Illinois.” EPA Solid Waste Management Series, SW—12d.

Haydar M.M. and Khire M.V. (2005a). “Leachate recirculation using horizontal trenches
in bioreactor landﬁlls.” Journal of Geotechnical & Geoenvironmental Engineering,
ASCE, Vol. 131, No. 7, 837-847.

Haydar, M.M. and Khire, M.V. (2005b). “Evaluation of anisotropy and heterogeneity of
waste hydraulic conductivity on leachate recirculation in MSW landﬁlls.” Journal of
Solid Waste Technology and Management, Vol. 30, No. 4, 233-243.

Haydar, M.M. and Khire M.V. (2005c), “Geotechnical Sensor system to Monitor
Migration of Injected Leachate in Municipal Solid Waste Landﬁlls.” Geotechnical
Testing Journal, ASTM, accepted.

Khire, M.V. and Haydar, M.M. (2003). “Numerical Evaluation of Granular Blankets for
Leachate Recirculation in MSW Landﬁlls.” Proceedings of the Ninth Sardinia Solid
Waste Conference, Cagliary, Italy, Oct.

Khire, M.V., and Haydar, M.M. (2005), "Leachate Recirculation Using Geocomposite

Drainage Layer in Engineered MSW Landﬁlls." Proceedings of the GeoFrontiers
2005, ASCE, Austin, TX, Jan 24-26.

Koemer, R. (2000). “Leachate Recycling Loading to Bioreactor Landﬁlls for the Rapid
Degradation of Municipal Solid Waste.” Proceedings of ESD Solid Waste
Management Conference, MI, 39 pgs.

Korﬁatis, G., Demetracopoulos, A., Bourodimos, E., and Nawy, E. (1984). “Moisture

transport in a Solid Waste Column.” Journal of Environmental Engineering, ASCE,
Vol. 110, No. 4, 789-796.

135

Maier, T. and Vasuki, N. (1997). “Expected Beneﬁts of a Full-Scale Bioreactor Landﬁll.
A Compilation of Current Readings: Leachate Recirculation & Landﬁll Bioreactors.”
SWANA, July 1997.

McCreanor, P. ( 1998). “Landﬁll leachate recirculation systems: mathematical modeling
and validation.” Ph.D. Dissertation, University of Central Florida, Orlando, FL.

Miller, D. and Emge, S. (1997). “Enhancing landﬁll leachate recirculation system
performance.” Practice Periodical of Hazardous, Toxic, and Radioactive Waste
Management, July, 1 13-1 19.

Mualem, Y. (1976). “A new model for predicting the hydraulic conductivity of
unsaturated porous media.” Water Resources Research, Vol. 12, 513-522.

Oweis, 1., Smith, D., Ellwood, R., and Greene, D. (1990). “Hydraulic characteristics of
municipal refuse.” Journal of Geotechnical Engineering, ASCE, Vol. 116, No. 4,
539-553.

Pang, L., Close, M., Watt, J ., and Vincent, K. (2000). “Simulation of picloram, atrazine,
and simazine leaching through two New Zealand soils and into groundwater using
HYDRUS-2D.” Journal of Contaminant Hydrol. , Vol. 44, No. 1, 19-46.

Pohland, F. and Kim, J. (1999). “In situ aerobic treatment of Leachate in Landﬁll
Bioreactors.” Water Sci. T echnol., Vol. 40, 203-210.

Qian, X., Koemer, R., and Gray, D. (2002). “Geotechnical Aspects of Landﬁll Design
and Construction.” Prentice Hall, New Jersey.

Rassam, D. and Cook, F. (2002). “Numerical simulations of water ﬂow and solute

transport applied to acid sulfate soils.” J. Irrig. Drain. Eng, Vol. 128, N0. 2, 107-
1 15.

Reinhart, D. and Al-Yousﬁ, B. (1996). “The impact of Leachate Recirculation on
Municipal Solid Waste Landﬁll Operating Characteristics.” Waste Management &
Research, Vol. 14, 337-346.

Reinhart, D. and Townsend, T. (1998). “Landﬁll Bioreactor Design & Operation”. Lewis
Publishers, Boca Raton, FL.

Simunek, J., Sejna, M., and Van Genuchten, M. (1999). “The HYDRUS-2D software
package for simulating the 2-D movement of water, heat, and multiple solutes in
variably Saturated media Version 2.0,” US. Salinity Laboratory, Agriculture
Research Service, USDA, Riverside, California.

136

Scanlon, B., Christman, M., Reedy, R., Form, 1., Simunek, J ., and F lerchinger, G. (2002).
“Intercode comparisons for simulating water balance of surﬁcial sediments in
semiarid regions.” Water Resources Research, Vol. 38, No. 12, 5901-5915.

Straub, W. and Lynch, D. (1982). “Models of landﬁll leaching: moisture ﬂow and
inorganic strength.” Journal of Environmental Engineering, ASCE, Vol. 128, 231-
250.

Sowers, G. (1972). “Settlement of Waste Disposal Fills.” Proceedings of 8th
International Conference on Soil Mechanics and Foundation Engineering, Moscow,
1972, 207-210.

van Genuchten, M.Th. (1980). “A closed-form equation for predicting the hydraulic
conductivity of unsaturated soils.” Soil Science Society of America Journal, Vol. 44,
892-898.

Wall, D. K. and Zeiss, C. (1995). “Municipal Landﬁll Biodegradation and Settlement,”
Journal of Environmental Engineering, Vol. 121, No. 3, 214-223.

Zeiss, C. and Ugguccioni, M. (1995). “Mechanisms and patterns of leachate ﬂow in
municipal solid waste landﬁlls.” Journal of Environmental Systems, Vol. 23, No.3,
247-270.

Zhao, X., Maher, 8., Khire, M., Musleh, R., Voice, T., and S. Hashsham (2004), “Field-

Scale Landﬁll Bioreactor Demonstration,” Proceedings of the S WANA Landﬁll
Symposium, Monterey, California, June.

137

PAPER NO. 5: GEOTECHNICAL SENSING SYSTEM TO MONITOR
INJECTED LIQUIDS IN LANDFILLS

ABSTRACT

A ﬁeld-scale horizontal permeable blanket made up of crushed recycled glass was
built in an active municipal solid waste (MSW) landﬁll to recirculate leachate. The
permeable blanket is a new method developed by the authors for subsurface leachate
recirculation or liquid injection. Leachate injection rates in the blanket ranged from
1.1 m3/m/hr to 3.6 m3/hr/m. An automated sensing system consisting of moisture
content sensors, pressure transducers, and temperature sensors was designed to
monitor the migration of injected leachate inside the blanket. The sensors were
embedded in the blanket and connected to a data logging system. All sensors were
able to detect the leachate migration within the blanket. The TDR and impedance
moisture content sensors could not detect the migration of injected leachate once the
surrounding medium got saturated. The pressure transducers and temperature sensors
were able to detect leachate migration irrespective of the degree of saturation of the
blanket. Unlike thermistor sensors, temperature readings measured by thermocouple

sensors were inﬂuenced by air temperature.

INTRODUCTION

Leachate recirculation has become a relatively common leachate management option for
MSW landﬁlls because it offers the following beneﬁts (Reinhart and Al-Yousﬁ 1996;
Miller and Emge 1997; Mehta et al. 2002; Haydar and Khire 2005): (1) a reduction in

leachate treatment and disposal costs; (2) accelerated decomposition and settlement of

138

waste resulting in a gain in airspace; (3) an increase in the rate of landﬁll gas production;
and (4) a potential reduction in the post-closure care period and maintenance costs. Risks
and drawbacks of leachate recirculation include: (1) a potential decrease in the factor of
safety for slope stability of landﬁlls; (2) a potential increase in the leachate head on the
liner if the leachate collection system is not designed to drain recirculated leachate; (3)
potential long-term clogging of the leachate recirculation and leachate collection systems;
(4) potential ﬂooding of the gas collection system; and (5) leachate seeps from the sides
of the landﬁll if an appropriate buffer between the leachate injection point and the sides
of the landﬁll is not maintained. Hence, before implementation, landﬁll operators should
weigh the advantages and risks associated with leachate recirculation on a site-speciﬁc
basis.

Subsurface leachate recirculation in MSW landﬁlls is commonly practiced using
conventional methods such as horizontal trenches and vertical injection wells. Khire and
Haydar (2003) have proposed the use of permeable blankets as an alternative for
subsurface leachate recirculation systems. Khire and Haydar (2003) have presented a
numerical study that compares hydraulic aspects of permeable blankets and horizontal
trenches and A ﬁeld-scale permeable blanket was constructed and instrumented at a
landﬁll located in Michigan to test the concept of leachate recirculation using permeable
blankets. Such testing was done using embedded sensors providing data on ﬂow and fate
of injected leachate in the blanket. The key objective of the ﬁeld-scale study presented
was to evaluate the feasibility of using an automated geotechnical sensing system to

monitor the migration of recirculated leachate in permeable blankets.

139

Leachate Recirculation Using Permeable Blankets

This design concept consists of a relatively thin and high hydraulic conductivity layer of
permeable material sandwiched between non-woven geotextile sheets laid on a relatively
ﬂat waste surface in a landﬁll. A schematic of a permeable blanket is presented in Figure
1. Thickness of such a blanket can vary depending upon the material used and site-
speciﬁc design and operational variables. A perforated pipe is embedded at the center of
such blanket in the direction parallel to the shorter or longer plan view dimension where
leachate is injected under a positive pressure. Installation of an injection pipe along the
short plan view direction would typically yield in ﬁnancial savings in the pipe related
costs. The areal dimensions and shape of the blanket can vary depending on the leachate
recirculation needs, shape of the landﬁll cell, relatiVe contrast in the hydraulic
conductivities of the blanket and the underlying waste, and leachate injection rate and
pressure. Khire and Haydar (2003) have presented speciﬁc details on the design variables
for permeable blankets used for leachate recirculation.

When leachate recirculation is practiced at MSW landﬁlls, it is important to
monitor migration of injected leachate due to the concerns associated with slope
instabilities of the landﬁll as a result of potential liquid head build up resulting in
reduction of waste shear strength (Qian et al. 2003). In addition, US. federal regulations

require that liquid head on the landﬁll lining system does not exceed 30 cm.

Common Sensors Used in Landﬁlls

The sensors most commonly used in landﬁlls can be grouped depending upon the

application as follows.

140

Embedded

' rs Sensors
Direction of a, may W l

(3)

Flow of Injected

  

------------

 

d Leachate

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.1 . waste as Injection Line
V
Permeable R, Undeﬂyng
Blanket Wetted
Am
Non-woven , Perforated Leachate
Geotextrle (GT) 1 Embedded Sensors Injection Pipe
1m '
GT Wrap Waste Permeable Drainage
(b) Around Material

 

 

 

 

 

 

Figure 1. Schematic of a permeable blanket (a) perspective view; and (b) cross section
AA’.

141

Moisture Content Sensors

Moisture content sensors commonly used in geotechnical applications include:

Time domain reﬂectometry (TDR): commonly used for agriculture and
environmental applications. Li and Zeiss (2000), among others, have investigated the
use of TDR to measure the volumetric water content of waste;

Time domain transmissometry (TDT): similar to TDR in operation and application;
Neutron probe: commonly used to estimate the moisture content of its surrounding
medium (Yuen et al. 2000) in shallow inorganic soils; and

Electrical resistance or impedance sensors: these sensors are either gypsum block
sensors (McCann et a1. 1992) which use resistance to direct current (DC) or
customized electrical impedance sensors (Gawande et al. 2003) which use resistance
to an alternating current (AC) to measure the moisture content of surrounding
medium. The gypsum block sensors do not function well in acidic or corrosive
environments of landﬁlls (Zhao et al. 2003). However, the impedance sensors
designed by Gawande et al. (Gawande et al. 2003) are not susceptible to corrosion
and hence perform better in landﬁlls. Impedance moisture content sensors are cheaper

compared to TDR or TDT sensors.

Temperature Sensors

Temperature sensors commonly used in geotechnical applications can be divided broadly

into thermocouple or thermistor sensors. Within the relatively narrow temperature range

expected in landﬁlls, thermistors are reported to be relatively stable and more accurate

142

compared to thermocouples (Reinhart et al. 2003). However, thermistor sensors are more

expensive.

Pressure Sensors

Monitoring wells and stand-pipe piezometers are conventional ﬁeld devices used to
monitor leachate pressure head or leachate level in landﬁlls. There are many landﬁlls
where pressure transducers are used to monitor leachate pressure. Reinhart et al. (2003)
have reported that pressure transducers can fail if appropriate precautions are not taken to
protect these sensors from lightening strikes and crushing or damage due to pressure of

waste compactor and overlying waste.

AUTOMATED SENSING SYSTEM

The required key characteristics of embedded sensors used for monitoring the migration
of injected leachate in permeable blankets in landﬁlls are: (l) the sensors are economical
and are durable in physically and chemically challenging environment of a landﬁll; and
(2) the sensor and the sensor system can be automated to make reliable measurements at
a frequency as high as once every few seconds. The automated sensing system used in
this ﬁeld-scale study considered all of the requirements presented above. The automated
sensing system consisted of the following sensors: (1) TDR sensors; (2) electrical
impedance moisture content sensors; (3) vibrating wire pressure transducers; and (4)

thermocouple and thermistor temperature sensors.

143

TDR Sensors

TDR sensors were chosen for this study because of their relatively well known history in

detecting moisture content of soils in the ﬁeld. The TDR sensor functions by emitting a

voltage signal. This signal is partially reﬂected when encountering a change in the

dielectric properties. From the travel time of the reﬂected signal and the length of the

TDR sensor, the bulk dielectric constant of the medium surrounding the TDR can be

determined. The bulk dielectric constant for permeable materials is the combined

dielectric constant of water, air, and solids. Because the dielectric constant of water (~

80) is much higher than that of air (~ 0) and soil solids (~ 2 to 4), the TDR sensor can

detect moisture content changes in surrounding medium. The TDR output readings can

be calibrated to obtain the volumetric water content of the surrounding medium.

Robinson et al. (1994), among others, have presented a detailed review on the progress

made in TDR sensor technology. Common factors that affect the TDR sensor

measurements in the ﬁeld are summarized blow:

0 organic content (McBean et al. 1994; Robinson et al. 1994);

0 presence of ferrous metal (Robinson et al. 1994; Dalton 1992);

0 saline soils or materials having higher electrical conductivity, which results in an
overestimation of moisture content (Herkelrath at al. 1991);

0 porosity. For a greater porosity, the TDR sensor readings increase due to greater
space for water to ﬁll in;

0 cable length. An increase in cable length decreases the magnitude of the reﬂected

signal and thus reduces the accuracy of the measurement (Robinson et a1. 2003); and

144

o variation in the ambient temperature of the cable. This might produce errors in

measurement especially when the cable is relatively long (2 35 m).

There are many commercial TDR sensors available for ﬁeld use (Roth et al. 1992).
The TDR sensor selected for this study consisted of a single rod (Figure 2) sealed in a
plastic casing. The reﬂected signal of the TDR sensor is translated into a current signal to
differentiate the small changes in the bulk dielectric constant. The output measurement
ranges from O to 1,000 mA for de-ionized (DI) water and increases as the electrical
conductivity (EC) of the surrounding liquid increases. The TDR sensor used in this study
is about 69-cm-long and has a diameter of 2 cm (Figure 2).

The use of electrical impedance concept to measure the moisture content of soils has
been in practice (e.g., gypsum block sensors) for many decades (Mch et al. 1992). A
photo of the electrical impedance moisture content sensor used in this study is presented
in Figure 2. This sensor, which was presented by Gawande et al. (2003), measures the
electrical impedance between two electrodes embedded in an uniformly graded sand

(average particle diameter, D50 = 1 mm) pack that is about 50 mm in diameter.

Electrical Impedance Moisture Content Sensors

The electrical impedance moisture content sensors have been commonly referred to in the
literature as electrical resistance moisture content sensors. However, because an

alternating current is used, these sensors measure impedance and not resistance.

145

 

Impedance moisture content
sensor with thennocouple
(Length = 24 cm and

g . 1 diameter = 7 cm)

. TDR sensor
“ .. ’ ' (Length 69 cm and
. ' . . p / diameter= 2 cm)

Pressure transducer
with thermistor
(Length =13 cm and
diameter = 1.9 cm)

 

Figure 2. Geotechnical sensors used in this study.

146

The selection of the impedance moisture content sensors to detect the leachate
migration in the permeable blanket was based on the following reasons: (1) the sensors
are durable in physically and chemically challenging environment of a landﬁll (Gawande
et al. 2003; Zhao et al. 2003); (2) good correlation exists between moisture content and
electrical impedance; and (3) the sensors are relatively cheap and unlike TDR sensors,
there is no limitation associated with the maximum cable length.

The electrical impedance measured between the electrodes is a function of the
moisture content of the sand pack and the surrounding medium. Liquids are absorbed into
the sand by capillary action enhanced by glass ﬁber wicks attached to the sensor body.
The impedance of the sensor is inversely proportional to the moisture content of the sand
pack. Liquid or leachate is absorbed into the sand from the surrounding material (waste)
or released depending on the matric potential difference between the sand pack and the
medium surrounding the sensor. The impedance moisture content sensor can capture the
changes in moisture content instantaneously. Gawande et al. (2003) reported that an
impedance value of 0.05 k!) or less represents that the sand pack is saturated. In order to
convert the impedance into moisture content of the surrounding medium, calibration
curves are required. A thermocouple of type T was added to the impedance moisture
content sensor to allow measurement of temperature.

Vibrating Wire Pressure Transducers

Vibrating wire pressure transducers measure the pressure head and serve as a substitute
for stand-pipe piezometers or monitoring wells. Unlike standpipe piezometers which can
clog up or interfere with daily landﬁll operations, the pressure transducers do not. If gas

pressure exists, these sensors measure combined gas and liquid pressure. When

147

barometric pressure is measured, vented pressure transducers can detect a change in the
pressure head irrespective of the moisture content. Pressure transducers cannot detect
moisture content. The pressure transducers made up of stainless steel are corrosion
resistant and advances in sealing technology have extended the life of these sensors in
challenging subsurface environments. The measurements from vibrating wire pressure
transducers have relatively low thermal sensitivity, are independent of the cable length,
and are relatively consistent and accurate (Duncliffe 1988). A thermistor is embedded
within the pressure transducer to record temperature and to correct the pressure
transducer measurement.

The pressure transducers use a pressure sensitive diaphragm attached to a vibrating
wire element. The diaphragm is welded to a capsule, which is evacuated and hermetically
sealed. When liquid pressure is exerted, the diaphragm is deﬂected and changes occur in
the tension and frequency of the vibrating wire. The changed frequency is sensed and
transmitted by electrical coil acting through the walls of the capsule to the readout device.
The pressure transducers selected for this study are not vented. Hence they required
correction for barometric pressure changes. This correction is signiﬁcant especially when
the pressure to be measured is relatively low.

Temperature Sensors

Because bacteria responsible for degradation of the waste and gas generation from a
landﬁll are most efﬁcient over a narrow temperature range, monitoring temperature
within a landﬁll can be useful in understanding or manipulating the operation of a landﬁll
to achieve an optimum performance. The role of temperature becomes even more

important when operating bioreactor landﬁlls (Tchbanoglous et al. 1993). The

148

temperature sensors selected in this study consisted of thermocouples and thermistors
which were built into the impedance moisture content sensors and the pressure
transducers, respectively. Because the temperature of the injected leachate is invariably
different compared to the background temperature of the blanket, these sensors could be
used to monitor the migration of the injected leachate within the blankets.

A thermocouple sensor consists of a junction of dissimilar metal wires. These wires
are of equal length and are soldered together. When both ends of the wires are at the
same temperature, there is no voltage generated. However, when the open ends are at a
ﬁxed temperature and the temperature of the junction changes, a voltage is generated
across the two wires. The temperature difference between the two ends of the wires is
correlated with the voltage difference. Hence, thermocouples require measuring the
reference temperature. The reference temperature is usually measured by a thermistor
located inside the datalogger box. The thermocouples used in this study were of type T
which is made of a positive copper wire and a negative constantan wire with a
temperature range of -200 °C to 350 °C.

Therrnistors are made of serni-conductor materials. The electrical resistance of
these materials changes with temperature. The resistance of the thermistor decreases as
the temperature increases. Unlike thermocouple sensors, thermistors measure the absolute

temperature so they do not require a reference temperature to make the measurements.

FIELD INSTRUMENTATION
A ﬁeld-scale horizontal permeable blanket made up of crushed recycled glass was
constructed at a municipal solid waste landﬁll located in Michigan to manage leachate by

leachate recirculation. This blanket is about 60-m-wide by 9.5-m-long. The average

149

thickness of the blanket is 150 mm. Installation of the blanket consisted of these steps in
the presented order: (I) laid a non-woven geotextile on a leveled waste surface of the
landﬁll cell; (2) placed crushed glass on the geotextile using a front end loader; (3)
installed sensors used for monitoring the migration of injected leacahte; and (4) covered
the glass blanket by a non-woven geotextile followed by placement of waste on the
blanket. The upper and lower geotextiles have a thickness of about 2 mm (ASTM D
5199), mass per unit of 270 g/m2, and a hydraulic conductivity of 0.3 cm/s (ASTM D
4491)

Hydraulic conductivities of the crushed glass were measured using a constant-head
test (ASTM D 2434-68) in loose and dense conditions (Table 1). Coefﬁcients of
uniformity and curvature (Cu and Cc), and the average particle size (D50) for the crushed
are equal to 1.4, 2.2 and 10 mm, respectively.

A 75-mm-diameter and 9.5-m-long perforated high-density polyethylene (HDPE)
leachate injection pipe was installed at the center of the blanket to inject leachate. The
end of the perforated pipe inside the blanket was capped and the other end was connected
to a pump to draw leachate from a series of interconnected above ground leachate storage
tanks having a total storage capacity of about 115 m3. A leachate ﬂow control valve, a
digital pressure gauge, and a magnetic ﬂow gauge were installed in the leachate pipe to

control and monitor the pressure head and ﬂow rate of injected leachate.

Monitoring System

The glass blanket was instrumented with embedded sensors consisting of TDR sensors,
impedance moisture content sensors with thermocouples, and vibrating wire vented

pressure transducers with thermistors. Barometric pressure sensor was also installed at

150

Table 1. Hydraulic conductivity of loose and dense crushed glass.

 

 

 

Property —> Dry Porosity Hydraulic
Density n Conductivity
State of Glass J, pd (gcm3) K (cm/s)
Loose 1.26 0.48 3.1
Dense 1.39 0.42 2.9

 

 

 

 

 

Note: Speciﬁc gravity of the glass G, = 2.41

151

the site to separate the barometric pressure component from the total pressure
measurement. Figure 2 shows a photo of these sensors and presents their dimensions.
Based on the topographic survey data and the dry weight of the crushed glass used for the
blanket, we estimated the ﬁeld porosity of the glass blanket to be about 0.45. The sensors
were installed at the bottom of the blanket just above the lower geotextile as presented in
Figure l(b). A vertical pressure sensor was installed just outside the boundary of the
blanket to monitor the vertical stress of waste overlying the glass blanket. Meteorological
sensors including a rain gauge, an air temperature sensor, and a barometric pressure
sensor were installed at the site. All sensors including the leachate ﬂow and pressure
gauge were connected to a data logger located at the site to continuously monitor and log
the data. The data logger was programmed to take readings at frequencies ranging from 1
minute to 1 hour.

Instrumentation of the blanket and sensing system took place in July 2003 followed
by ﬁlling of about 3-m-thick waste above the blanket in Sep. 2003. The thickness of
waste placed above the blanket before leachate recirculation trials began was greater than
the maximum leachate injection pressure head to prevent potential artesian conditions.
Over 80 leachate recirculation trials have been conducted since Sep. 2003 at leachate
injection rates ranging from 1.1 to 3.6 m3 per hour per meter length of the injection pipe
(m3/hr/m). Over 3,000 m3 of leachate has been recirculated in the landﬁll through the

glass blanket since Sep.2003.

152

Laboratory Testing and Calibration

Before installation in the ﬁeld, the instrumented sensors were tested in the lab to evaluate
the accuracy of their measurements and to develop calibration curves speciﬁc to the
landﬁll leachate.

In order to check the accuracy of the pressure transducer, the sensor was embedded
in a container containing crushed glass and tested using solutions of DI water and
potassium chloride (KCl) at liquid levels ranging from 0 to 70 cm. We observed a linear
relationship between the depth of water in the container and the pressure head readings
measured by the pressure transducer and it was independent of the EC of the liquid. The
accuracy of the pressure transducer was within :1: 2 cm over the entire measurement range
of the transducer.

The TDR and impedance moisture content sensors were fully submerged in a solution
of KCl and DI water having EC ranging from 0 (pure DI water) to 20 mS/cm. Figure 3
shows the response of the TDR and impedance moisture content sensors as a function of
the EC of the liquid medium. This response was obtained by fully submerging the sensor
in liquids having EC presented on the x-axis in Figure 3. The hatched area in Figure 3
shows a range of 5 mS/cm to 10 mS/cm for the EC of the leachate from the landﬁll. For
this range of BC, the impedance moisture content sensor readings (R) at saturation ranged
from 0.025 k!) to 0.045 k0 whereas the TDR sensor readings ranged from 1,320 to 1,380
11A. The impedance moisture content sensor readings were about the same when
submerged in KC] solution with or without crushed glass for a given EC. However, as
presented in Figure 3, the TDR sensor readings dropped slightly (e.g., drop from 1320 to

1230 11A for EC = 5 mS/cm) when submerged in the KCl solution with

153

.8

 

 

 
  
  

TDR sensor fully submerg 1
in saline solution

a -- - 'TDR sensor submerged """
' in crushed glass saline solution

_J

Range of electrical conductivity 1
- for leachate from McGill Landﬁll .

Impedance moisture content
sensor fully submerged
in saline solution

 
 
 

TDR Moisture Content Sensor Reading (11A)

De' ' d W§
ionize a
1000 m4 . L 1 m 1 r - r 1 r 1 1 1
0 5 10 1 5 20
Electrical Conductivity of KCI Solution (mSIcm)

O
mpedance Moisture Content Sensor Reading, R(kn)

 

0.01

Figure 3. Lab evaluation of electrical conductivity of liquid on the measurements by TDR
and impedance moisture content sensors.

154

crushed glass having a porosity of about 0.45. The drop in the TDR reading is because
the glass particles reduced the bulk dielectric constant of the medium.

The rate of drainage of the impedance moisture content sensors is a function of the
rate of drainage of the sand pack in these sensors. The rate of drainage of the sand pack is
a function of the hydraulic properties of the sand as well as the material surrounding
(crushed glass) and underlying (waste) the sensor (Figure 1b). Because the hydraulic
conductivity of the sand pack is almost three orders of magnitude less than the
surrounding crushed glass blanket, after a wetting event, the sensors drained slower than
the surrounding glass did. The rate of wetting and drying (draining) of the sensor is also a
function of hysteresis in the soil-water characteristic properties of the sand pack.

The TDR moisture content sensor is located in a sealed plastic casing. The rate of
drainage of the TDR sensor is directly proportional to the bulk dielectric constant or the

moisture content of its surrounding medium located within a 2-cm-radious.

RESULTS

Detection of Leachate Migration in Permeable Blankets

Response of T DR and Impedance Moisture Content Sensors

Figure 4 shows the response of ﬁve impedance moisture content sensors located at 4.5 to
22.5 m distance from the leachate injection pipe and two TDR sensors located at 9 and 18
m distance. This response corresponds to a leachate injection event at leachate injection
rate equal to 1.1 m3/hr/m. The positive values on the x-axis (0 to 300 minutes) represent

the time since the leachate injection began. The negative values on the x-axis (-100

155

minutes to 0) represent the time before the leachate injection was started. The two arrows
on the x-axis correspond to when the leachate injection was started and turned off.
During this event, about 45 m3 of leachate was injected in the glass blanket in 270
minutes. The TDR moisture content sensor readings immediately before the leachate
injection were about 300 11A and the impedance moisture content sensor readings ranged
from 0.03 k0 to 0.13 k9. This wide range of initial readings of the impedance moisture
content sensors reﬂects the spatial variation in the initial water content of the blanket.
The reason for this variation in the initial water content of the blanket can be attributed to
many reasons including: (1) low elevation pockets in the glass blanket; and (2) spatial
variation in the hydraulic properties and water content of the underlying waste. Based on
the lab measurements, the TDR sensor reading of 300 uA and impedance moisture
content sensor reading of 0.13 k!) indicated partial saturation whereas the impedance
moisture content sensor reading of 0.02 kﬂ indicated saturation. Even though these
sensors are located relatively close to each other, the impedance sensor indicated much
higher saturation and in some cases it saturated faster (Figure 4) because the impedance
sensor is measuring the saturation of the sand pack inside the sensor which has a greater
capillary suction. Because the hydraulic conductivity of the sand pack (~ 10’3 cm/s) is
three orders of magnitude less than that of the surrounding crushed glass (~ 1 cm/s), the
sand pack does not drain as fast as the surrounding crushed glass.

After the leachate injection began, all impedance moisture content sensors except
the one located at 22.5 m distance indicated a decrease in the impedance to about 0.03
k0. Such a decrease in the impedance occurred progressively with time with the sensor at

4.5 m showing the decrease within 20 minutes after the injection began and the sensor at

156

 

. . . - as. ... - . . . . . . . , . 1400
Leachate Injection Rate = 1.1 m3/m/hr

.
O
...L

 

 

 

Moisture Content Sensor Impedance, R (kn)
TDR Moisture Sensor Reading ( 11A)

 

 

0 1 T 9“ 9:1@
g=i;1;1§‘.§iélél515.51.35.155 , . , , . . z " ‘ Leaca hte Injection ~ 600
Duration = 0-270 min:
’ TDR @ TDR @ i
9 m 18 m a 400
0.5 1 1 1 J 1 1 1 1 l 1 I 1 L 1 1 L ‘+' q 200
-100 0 100 200 300

Time (minutes)

Figure. 4. Field response of TDR and impedance moisture content sensors to a leachate
injection event.

157

18 m showing the decrease at about 150 minutes. This response is due to the arrival of
the leachate front of the injected leachate. The impedance sensor located at 22.5 m
distance was saturated before the leachate injection began. Hence, this sensor did not
show any reduction in its impedance in response to the arrival of the injected leachate.
The TDR sensors located at 9 and 18 m distance showed increase in the readings from
300 to about 1,200 uA within 10 minutes after the impedance sensors at proximity

locations responded to the arrival of injected leachate.

Response of Pressure Transducers and Thermistosr

Figure 5 shows the response of the thermistor and pressure transducer sensors embedded
in the glass blanket at 0.5, 4.5, and 14 m distances from the leachate injection pipe. After
the leachate injection began, all pressure transducers indicated a gradual increase in the
pressure head (hp) in response to the leachate injection event. The increase in the pressure
head was earliest and greatest for the sensor located closest (at 0.5 m) to the leachate
injection pipe. After the leachate injection stopped at about 90 minutes, all pressure
transducers indicated an immediate but gradual decrease in the leachate pressure head.
For a speciﬁc leachate injection event, we believe that the increase in pressure
heads recorded by the pressure transducer sensors was due to liquid pressure increase.
This increase was not gas pressure buildup in the landﬁll because: (1) the response of all
three pressure transducers was synchronized with respect to the leachate injection event;
and (2) the response of thermistors in the pressure transducer sensors and of neighboring

moisture content sensors was also concomitant to the leachate injection event.

158

 

70.e._ T.s.24

    
    
 
   
 
   

 

 

 

 

_ f T f f Lachale injection Duraiion: 0 -90 min .

60 .1 " g 9 ° - e ’1 Leachate Temperature = 14.5 °C1
E "T @ 14 mm = 1
3 1 22
.cn 50 ~-
13“ 1 Leachate Injeaction _ 5';
0 40 T Rate = 3.6 m /m/hr .. 20 ..
I @ l-
o h . 6
I. p A .-
3 30 0.5 m ‘ 3
s ,8 s
E 20 4 a
2 s
g 10 i-
s: 16
o
-’ o

-10 . 1 ‘ . ‘ 1 . ' 1 ' 1 . . 1 14

-1 00 0 100 200 300

Time (minutes)

Figure 5. Field response of pressure transducers and thermistors to a leachate injection
event.

159

The temperature of the injected leachate was about 14.5 °C for the event presented
in Figure 5. In response to the leachate injection event, all three thermistor sensors built-
in with the three pressure transducers indicated a slight decrease in the temperature (T).
This decrease in the temperature correlates well with the increase in the pressure heads
measured by the sensors. Figure 5 indicates that thermistors are capable of measuring
change in temperature induced by the migration of the injected leachate.

Data presented in Figures 5 and 6 conﬁrmed that injected leachate migrated
horizontally in the permeable blanket due to relatively high hydraulic conductivity of the
crushed glass in the permeable blanket. This ﬁeld data veriﬁes the key ﬁnding of the
numerical evaluation of permeable blankets presented by Khire and Haydar (2003) and

Haydar and Khire (2005b).

Limitations of Sensor System to Detect Leachate Migration

Figure 6 shows the response of two impedance moisture content sensors and the three
pressure transducers for a leachate injection event at leachate injection rate equal to 1.1
m3/hr/m that occurred on 17 Dec. 2003. The impedance moisture content sensor readings
of about 0.03 k0 prior to the leachate injection event indicated saturation of the
surrounding medium. Hence, the impedance moisture content sensors did not show an
explicit response to the subsequent leachate injection event. Similarly, the TDR sensors
did not show an explicit response to the leachate injection event (not shown in Figure 6).
However, as presented in Figure 6, all three pressure transducers successfully detected
leachate migration during the leachate injection event by measuring a sharp increase in

the leachate pressure head followed by a gradual drop once leachate injection was

160

 

    
     
    

 

 

 

 

50 "- T" " F—T'MHM "WTMF—T _1' _‘ "T—""'" ”T' 'T‘TT’Y“_" ' ~7"'"‘_“" T T I T 1 .T 1 a
g 17 Dec. 200 ' . . i x
1 Leachate Injection : E
A 1' Rate = 1.1 m3/m/hr . .
E 40 -- a
3 : Leachate Injection ‘ g
1:“ - Duration: 0 - 140 min in
. r h @ 0.5 m '2
'o 30 — P 1:.
'° ” E
o L- -
I L I.
2 ~ .9.
3 2° 1" o 1 c
g j h @ 4 5 m , N", 8
E 1 +-
g_ - c
a, 10 _. h @ 14 m - , 3
*5 P: 3,, . e 3‘ " p I ~ 9 cabin==uiu g
f, ””3“” .1‘ «'1' Halt-1 ~43"). s 1 @1111 I 1311', 0
E 0 nesudne=au - i=3“ * ' ' so... ° e 0 : : 9.9 3,4,», 9% x g
I 4 R @ 4.5 and 14 m E
¢:::°:3:¢6:' * g

_1 0 1 1_ 1 1 1 g . 1 1 1 1 m 1 7 0.01

-120 O 120 240 360 480

Time (minutes)

Figure 6. Field response of pressure transducers and impedance moisture content sensors
to a leachate injection event. -

161

stOpped. Thus, the pressure transducers were capable of measuring increases in the
leachate pressure head regardless of the initial state of saturation of the glass blanket.

This event illustrates that moisture content sensors can detect the arrival or
migration of leachate only if the surrounding medium is not saturated. Once the medium
gets saturated, the sensor system design needs to have redundancy built in by having
pressure transducers and/or temperature sensors to detect the migration of injected
leachate.

The response of temperature sensors to the migration of injected leachate depends
on the temperature difference between the injected leachate and the temperature of the
blanket prior to leachate injection, rate of injection and quantity of injected leachate,
distance between the injection point and the sensor location, and thermal properties of the
permeable blanket. Most often, a temperature sensor will detect the injected leachate as
long as there is sufﬁcient difference between the temperature of the leachate and the

blanket.

Drainage of Permeable Blanket

The key function of permeable blankets is to drain or inﬁltrate injected leachate into the
underlying waste and hence wet the waste. The rate of drainage of the injected leachate
through the blanket is a function of the hydraulic properties of the blanket as well as the
underlying waste and the volume and frequency of leachate dosing. Immediately after the
completion of a typical leachate dosing or leachate injection event, the blanket partially
drains. During the drainage of the blanket, highest moisture content within the blanket

resides at the bottom, where all embedded sensors are located.

162

To test the drainage of the blanket, leachate injection was paused during the period
from 15 Oct. to 23 Nov. 2003. The TDR sensors responded to the drainage of the blanket
sooner than the impedance moisture content sensors. During this period, the TDR
readings decreased from about 1200 to 200 uA. While, the impedance measured by the
moisture content sensors increased by an order of magnitude from about 0.03 to 0.3 kﬂ
during the same period. This change in readings of the TDR and impedance moisture
content sensors reﬂected the drying or drainage of the permeable blanket.

Drying of these sensors was interrupted by an increase in the moisture content of
the glass blanket due to precipitation that occurred on 25 Oct. 2003 and 4 Nov. 2003

when a total 20 mm precipitation was recorded.

Effect of Diurnal Air Temperature on Measurements

The thermocouple sensors used in this study are built within the impedance moisture
content sensors. The thermocouple junction is located between the sand pack and the
PVC casing of the sensor (Reinhart et al. 2003). Hence, the thermocouple temperature
readings can be impacted by relatively small changes in temperature that may occur due
to the absorption and release of leachate from the sand pack. Throughout the monitoring
period, the thermocouple readings were noisy and fluctuated regularly by about 2 °C.
Figure 7 shows the response of thermocouple and thermistor sensors located in the
blanket on 8 Nov. 2003. The air temperature and the reference temperature measured by
the datalogger are also presented in Figure 7. On 8 Nov. 2003, leachate was not injected
in the blanket and no precipitation was recorded. Figure 7 indicates that the temperature
readings measured by all thermistors were constant during the entire day. However, the

temperature readings measured by all thermocouples showed a gradual increase followed

163

 

  
 

1 I l
Increase in thermocouple; '
temperature readings;

     
  

 
 
  
 
   
     

 

 

 

co smsussssmssm::'a::m:m::§=:
:- 20 F Relatively steady thennistorE/ i
s s temperature readings; Reference .
E t I Temperature 1
a 10 ~— ' —
s » -
l- : :
P 4
O — _
~\ . Air : -
: z i Temperature § q
-10 _ j i i
0:00 6:00 12:00 18:00 23:59

8 November 2003 - Time of Day

Figure 7. Effect of diurnal air temperature on the ﬁeld measurements of thermistors and
thermocouples.

164

by a gradual decrease in tune with the air temperature between 8 AM to 6 PM. The
change in the temperature recorded by the thermocouple sensors is due to the changes in
air temperature affecting the temperature of the thermocouple cable and hence the
temperature of the reference thermistor junction. Because a typical thermistor sensor is
insulated from its cable, the thermistor readings were not impacted by the changes in the

air temperature.

SUMMARY AND CONCLUSIONS

Migration of injected leachate in permeable blankets can be appropriately monitored
using an automated sensing system consisting of moisture content, temperature, and
pressure sensors. Pressure transducers and thermistors were able to detect the migration
of injected leachate in the permeable glass blanket irrespective of the initial degree of
saturation of the blanket. TDR and impedance moisture content sensors detected injected
leachate only when the surrounding medium was not saturated prior to leachate injection.
Even though moisture content sensors will not be able to detect leachate migration once
saturated, these sensors are important elements of the monitoring system because these
sensors provide the state of saturation of the surrounding medium and provide
redundancy in the measurement when unsaturated.

Temperature measurements by thermocouple sensors were affected by the diurnal
ﬂuctuations in air temperature affecting the reference temperature. Temperature
measurements by thermistor sensors were not affected by the diurnal ﬂuctuations in air
temperature. Hence, thermistors are more reliable to measure temperature and to detect

leachate migration compared to thermocouples. We recommend the use of an automated

165

sensing system consisting of moisture content as well as temperature and pressure

transducer sensors to provide redundancy in the measurements.

ACKNOWLEDGMENTS

The ﬁeld-scale testing was jointly funded by Waste Management Inc. and Environmental
Research & Education Foundation (EREF). We sincerely appreciate the support from the
sponsors. The authors would also like to express sincere appreciation to Ronald “Hog”
Feldkamp, Chester Stanley, and Paul Mazanec for coordination of numerous leachate
recirculation trials at the McGill Landﬁll. We also appreciate the assistance provided by
Dr. Philip McCreanor related to the impedance moisture content sensors. The results and
opinions presented in this manuscript are those of the authors and have not been reviewed

by anyone else.

166

REFERENCES

Dalton, F. N., 1992, “Development of Time-domain Reﬂectometry for Measuring Soil
Water Content and Bulk Soil Electrical Conductivity,” In: Topp, G. C., W. D.
Reynolds, and R. E. Green (Eds), Advance in Measurement of Soil Physical
Properties: Bringing Theory into Practice, Soil Science Society of America,
Special Publication No. 30, pp. 143-167.

Duncliffe, J., 1988, Geotechnical Instrumentation for Monitoring Field Performance.
John Wiley, New York.

Gawande, N. A., D. R. Reinhart, P. A. Thomas, P. T. McCreanor, and T. G. Townsend,
2003, “Municipal Solid Waste in Situ Moisture Content Measurement Using an
Electrical Resistance Sensor,” Waste Management, Vol. 23, No. 7, pp. 667-674.

Haydar M. M. and M. V. Khire, 2005, “Leachate recirculation using horizontal trenches
in bioreactor landﬁlls,” Journal of Geotechnical & Geoenvironmental Engineering
Journal, ASCE, in press.

Herkelrath, W. N., S. P. Hamburg, and F. Murphy, 1991, “Automatic Real-time
Monitoring of Soil Moisture in a Remote Field Area with Time Domain
Reﬂectometry,” Water Resources Research, Vol. 27, No. 5, pp. 857-864.

Khire, M. V. and M. M. Haydar, Oct. 2003, “Numerical Evaluation of Granular Blankets
for Leachate Recirculation in MSW Landﬁlls,” Proceedings of the Ninth Sardinia
Solid Waste Conference, Cagliary, Italy.

Li, R. S. and C. Zeiss, 2000, “In situ Moisture Content Measurement in MSW Landﬁlls
with TDR,” Environmental Engineering Science, Vol. 18, No. 1, pp. 53-66.

McBean, E. A., F. A. Rovers, and GJ. Farquhar, 1994, Solid Waste Landﬁll Engineering
Design. Pearson Education, New York.

McCann, I. R., D. C. Kincaid, and D. Wang, 1992, “Operational Characteristics of the
Watermark Model 200 Soil Water Potential Sensor for Irrigation Management,”
Applied Engineering in Agriculture, Vol. 8, No. 5, pp. 603-609.

Mehta, R., Barlaz, M., Yazdani, R., Augenstein, D., Bryars, M., and Sinderson, L., 2002,
“Refuse Decomposition in the Presence and Absence of Leachate Recirculation,”
Journal of Environmental Engineering, Vol. 128, No. 3, pp. 228-236.

Miller, D. and Emge, 8., July 1997, “Enhancing landﬁll leachate recirculation system

performance,” Practice Periodical of Hazardous, Toxic, and Radioactive Waste
Management, pp. 113-119.

167

Qian, X., R. M. Koemer, D. H. Gray, 2003, “Translational Failure Analysis of Landﬁlls,”
Journal of Geotechnical and Geoenvironmental Engineering, ASCE, Vol. 129, No.
6, pp. 506 — 519.

Reinhart, D. R., T. Townsend, and P. McCreanor, 2003, “Instrumentation for In Situ
Monitoring of Municipal Solid Waste Landﬁll Processes,” Proceedings of the Ninth
Sardinia Solid Waste Conference, Cagliary, Italy, Oct.

Reinhart D. and Al-Yousﬁ B., 1996, “The impact of Leachate Recirculation on Municipal
Solid Waste Landﬁll Operating Characteristics,” Waste Management & Research,
Vol. 14, pp. 337-346.

Robinson, D. A., J. P. Bell, and C. H. Batchelor, 1994, “The Inﬂuence of Iron Minerals
on the Determination of Soil Water Content Using Dielectric Techniques,” Journal
of Hydrology, Vol. 161, No. 1-4, pp. 169-181.

Robinson, D. A., S. B. Jones, J. M. Wraith, D., and S. P. Friedman, 2003, “A Review of
Advances in Dielectric and Electrical Conductivity Measurement in Soils Using
Time Domain Reﬂectometry” Vadose Zone Journal, Vol. 2, No. 4, pp. 444-475.

Roth, C. H., M. A. Malicki, and R. Plagge, 1992, “Empirical Evaluation of the
Relationship between Soil Dielectric Constant and Volumetric Water Content as the

Basis for Calibrating Soil Moisture Measurements,” Journal of Soil Science, Vol.
43, No.1, pp. 1-13.

Tchobanoglous G., H. Theisen, and S. Vigil, 1993, Integrated Solid Waste Management.
McGraw-Hill, New York.

Yuen, S. T. S., T. A. McMahon, and J. R. Styles, 2000, “Monitoring In situ Moisture
Content of Municipal Solid Waste Landﬁlls,” Journal of Environmental
Engineering, Vol. 126, No. 12, pp. 1088-1095.

Zhao, X., S. Maher, R. Musleh, M. Khire, T. Voice, and S. Hashsham, June 2003, “Full-

scale Evaluation of Bioreactor Landﬁll Technology,” Proceedings of SWANA 8'”
Annual Landﬁll Symposium, Atlantic City, NJ.

168

PAPER NO. 6: LEACHATE RECICRCULATION IN BIOREACTOR
LANDFILLS USING GEOCOMPOSITE DRAINAGE MATERIAL

ABSTRACT

The key purpose of this study was to test the use of a permeable blanket made up of
geocomposite drainage layer (GDL) for leachate recirculation in municipal solid waste
(MSW) landﬁlls. A 34-m-long by 12-m-wide permeable blanket made of geocomposite
drainage layer (GDL) was constructed at an active MSW landﬁll located in Michigan.
Leachate was injected in the GDL using a perforated pipe placed centrally above the
GDL along its length. Moisture content sensors, pressure transducer, thermistor,
thermocouple sensors, and a vertical load sensor were embedded immediately below the
GDL blanket to monitor the ﬂow of injected leachate. Afterthe blanket was covered with
waste, leachate was injected into the blanket at rates ranging from 0.9 to 2.6 m3/hr per
meter length of the blanket. Data collected from the embedded sensors indicated that the
injected leachate traveled at rates ranging from 5 to 18 m/hr through the blanket
depending upon the leachate injection rate. The rate of travel of injected leachate through
the GDL blanket was not uniform. Only pressure transducers and thermistors were able to
detect migration of injected leachate once the blanket got saturated and moisture content
sensors could not register any change in readings. Leachate injection pressure monitored
over a period of two years indicated no signs of clogging of the blanket. The leachate
pressures measured immediately below the blanket were less than the net leachate
injection pressure indicting that there was a head loss in the GDL blanket. Based on this
ﬁeld data, we conclude that permeable blanket made of GDL can be used to recirculate
leachate in MSW landﬁlls. However, long term monitoring of the performance of the

blanket needs to be evaluated.

169

BACKGROUND

Environmental and economical beneﬁts of leachate recirculation in municipal solid waste
(MSW) landﬁlls are well documented (Reinhart and Al-Yousﬁ 1996; Reinhart and
Townsend 1998; Pohland and Kim 1999). These beneﬁts include: (1) reduction in
leachate treatment and disposal costs; (2) accelerated decomposition and settlement of
waste resulting in an airspace gain; (3) an increase in the rate of gas production; and (4)
potential reduction in the post-closure care period and maintenance costs. Risks
associated with leachate recirculation include: (1) potential decrease in the factor of safety
for slope stability of the landﬁll; (2) potential increase in the leachate head on the liner if
the leachate collection system is not designed to drain recirculated leachate; (3) potential
ﬂooding of the gas collection system; and (4) leachate seeps from the sides of the landﬁll
if an appropriate buffer distance is not maintained. Before implementation of a leachate
recirculation system (LRS), designer and landﬁll operator are expected to weigh the
advantages and disadvantages on a site-speciﬁc basis.

The most common leachate recirculation techniques are broadly divided into
surface and subsurface application. Surface applications consist of: (1) the direct
application of leachate or spray irrigation of leachate on the landﬁll surface; and (2)
surface ponding of leachate. Conventional subsurface application methods are: (1)
vertical wells; and (2) horizontal trenches (Koemer 2000, Qian et al. 2002; Haydar and
Khire 2005a).. Horizontal trenches (HTs) are commonly used in relatively new landﬁlls,
while vertical wells are mostly used in retroﬁt landﬁlls or where implementing HTs is not
feasible or cost effective. Haydar and Khire (2005a) showed that for an LRS consisting of

HTs, horizontal spacing of HTs needs to be less than 10 m for most commonly used

170

injection pressure heads ( up to 5 m) to reduce the formation of dry zones where injected
leachate cannot reach.. Khire and Haydar (2003) and Khire and Haydar(2005b) presented
a numerical study that proposed the use of a new subsurface LRS called “permeable
blankets” as an alternative to the conventional systems. The key advantages of permeable
blankets (PBS) over conventional systems are: (1) excavation of waste is not needed
during the construction of the blanket resulting in fewer odor problems; (2) a relatively
uniform distribution of leachate in the landﬁll potentially reducing differential settlement
and resulting post-closure maintenance costs; and (3) a PB can substitute multiple

horizontal trenches or vertical wells and can offer better cost to overall beneﬁt ratio.

Leachate Recirculation Using Permeable Blankets

This design concept consists of laying a relatively thin layer of permeable material (e.g.,
shredded tires, crushed glass, pea gravel, GDL, etc.) having relatively high hydraulic
conductivity on a ﬂat or inclined waste surface in a landﬁll (Khire and Haydar 2003). A
perforated pipe is embedded in the blanket in the direction parallel to the shorter or longer
plan view dimension of the blanket where leachate is injected under a positive pressure.
The thickness of such a blanket can vary depending upon the material used and the
site-speciﬁc design and operational variables. The aerial dimensions and the shape of the
blanket can vary depending on the leachate recirculation needs, the shape of the landﬁll
cell, relative contrast in the hydraulic conductivities of the blanket and the underlying
waste, and leachate injection pressure or leachate injection rate. Haydar (2005) has
presented relevant details on design and performance of leachate recirculation using

permeable blankets in MSW landﬁlls.

171

There are key concerns that need to be considered when designing a LRS using
PBs. Biological activities that occur in landﬁlls may partially or fully clog the permeable
blanket and decrease its hydraulic conductivity and hence may jeopardize the hydraulic
efﬁciency of the LRS. When a blanket settles or sags, a higher leachate injection head is
needed to compensate for the elevation head loss. If permeable blankets are subjected to
relatively high non-uniform vertical loads, differential settlement may jeopardize the
physical integrity and the hydraulic continuity of the blanket and can cause disruption in
the ﬂow within the blanket. When using permeable blankets, a daily cover with
permeability greater than that of waste should be used to minimize excessive lateral
spreading of injected leachate and allow vertical spreading (McCreanor 1998).

Field-scale data on performance of leachate recirculation systems is relatively
scarce. Hence, designing a leachate recirculation system to achieve a target performance
is a challenge for landﬁll designers. Data from an appropriately designed ﬁeld-scale
monitoring system would greatly beneﬁt landﬁll designers. Such data is especially useful
because there are potential risks including the physical stability of landﬁll that must be

quantiﬁed on a site-speciﬁc basis before an LRS is designed and implemented.

OBJECTIVES

In this ﬁeld-scale study, a permeable blanket made up of geocomposite drainage material
was constructed and instrumented at an active MSW landﬁll. The key objectives of the
ﬁeld-scale study were to: (1) test the use of geocomposite drainage layer (GDL) as a
permeable blanket to recirculate leachate in municipal solid waste (MSW) landﬁlls; (2)

evaluate the distribution of injected leachate in the GDL blanket; and (3) monitor the

172

leachate injection pressure, pressure head buildup in the blanket, and temperature change

of leachate in the blanket.

GEOCOMPOSITE DRAINAGE LAYER

A GDL is a ﬂat geosynthetic product that is used in landﬁll as well other civil and
environmental drainage applications. GDLS in landﬁlls have been primarily used as
leachate collection layers, leak detection layer between primary and secondary liners of
double lined landﬁlls, and lateral drainage layer to drain inﬁltrated precipitation in
landﬁll caps (Koemer 1999). A GDL is made up of a geonet (GN) layer sandwiched
between geotextiles (GT) or geomembranes. Its purpose is to achieve relatively high
transmissivity to maximize the lateral ﬂow through the geonet. A S-mm-thick geonet
layer can be hydraulically equivalent to about a 200- to 300-mm-thick gravel drainage
layer under equivalent hydraulic conditions. Most commonly, in a typical GDL, non-
woven geotextiles are attached to a geonet on either sides. The key purpose of geotextiles
is to separate the GN from the surrounding porous material (e.g., soil, waste, etc.) and to
prevent clogging of the GN.

Eith and Koemer (1992) have summarized these advantages of geonets or GDLs
over natural soils for landﬁll applications: (1) saving of valuable landﬁll air space
because these products are relatively thin; (2) relatively simple construction as it does not
require heavy equipment for placement and compaction; (3) cost savings depending on
the proximity of good-quality natural drainage material as an alternative; and (4) possible
reduction or elimination of perforated pipes to drain liquids from the system.

When compared to other permeable materials, blanket made up of GDL has these

key advantages: (1) better physical integrity against differential loading or settlement;

173

and (2) relatively small thickness. These key factors must be considered when using
permeable blankets made of GDL: (1) potential clogging of the geotextiles used in GDLS
due to bacterial grth or scale formation due to the precipitation of minerals from
leachate; (2) temperature of the leachate potentially impacting the physical properties of
the GDL; and (3) creep under relatively large conﬁning stresses (Giroud et al. 2000).
Koemer et al. (1994) found that leachate containing greater than 2,500 g/ml of total
suspended solids or biochemical oxygen demand needs laboratory simulation to assess

the clogging effect on a site-speciﬁc basis.

NUMERICAL MODELING OF PERMEABLE BLANKETS
In order to illustrate the conceptual hydraulics of permeable blanket and how it compares
to the conventional HTs, we have presented the wetted areas at steady-state simulated for
a horizontal permeable blanket and horizontal trenches in Figure 1. The simulated wetted
area for a permeable blanket placed on an inclined surface is presented in Figure l(b).
These wetted areas are simulated by Haydar and Khire (2005a, 2005b) by using the
saturated/unsaturated ﬂow model HYDRUS-2D developed by Simunek et a1. 1999. The
waste was assumed homogeneous and isotropic for simulating the wetted areas presented
in Figure 1. Because waste is a highly heterogeneous material, this assumption is
simplistic. However, it allowed us to compare the two designs. In the later part of this
manuscript, we have addressed the effect of waste hydraulic conductivity on the rate of
travel of leachate in the blanket.

Haydar and Khire (2005b) also have presented a numerical parametric study for
horizontal permeable blankets for leachate recirculation in MSW landﬁlls. The objective

of the parametric study presented by Haydar and Khire (2005b) was to evaluate the effect

174

HTs Spaced at 15 m Wetted Area
spacing 7 \ / I (2 90% Saturation)

 

(a)

 

 

 

Horizontal Permeable
Blanket —\ [— Perforated Leachate Injection Pipe

 

 

l l

(b) \

\—— Wetted Area
(2 90% Saturation)

 

 

 

 

/— Perforated Leachate
Injection Pipe

Sloped Permeable
Blanket

\— Wetted Area
(2 90% Saturation)

(C)

 

 

Figure 1. Schematic of wetted area at steady-state simulated by HYDRUS-2D for
leachate recirculation system consisting of: (a) four horizontal trenches; (b) a horizontal
permeable blanket; and (c) a sloped permeable blanket.

175

of key design parameters on the rate of travel and the pressure head of injected leachate
in the blanket. The authors deﬁned the wetted width as the maximum distance traveled by
the injected leachate at steady-state or during a ﬁnite duration measured from the
injection pipe in the blanket. The wetted width can be used to design the length of the
blanket and to determine an appropriate buffer distance to the landﬁll side slopes to
minimize leachate breakouts. The blanket width inﬂuences the liquid pressure head in the
blanket. Haydar and Khire (2005b) showed that wetted width is primarily a function of
the leachate injection rate, dosing frequency of injected leachate, and hydraulic
conductivity of underlying waste (kw). Assuming all other parameters constant, an
increase in the liquid injection rate or dosing frequency, or a decrease in kW results in a
greater wetted width.

The thickness or depth of permeable blanket has very little effect on the wetted
width. However, increase in the PB depth increase the storage capacity of the PB and
hence reduces liquid pressure buildup. Thus, under equivalent conditions, a thicker PB
would yield a greater factor of safety against static slope stability. Haydar and Khire
(2005b) also found that decrease in the hydraulic conductivity of the blanket (k3) requires
an increase in the leachate injection pressure head (larger head pump) to maintain the
injection rate constant and results in a greater liquid pressure buildup in the blanket.

The ﬁeld-scale study presented in this manuscript was designed to verify the PB
concept and to monitor the liquid injection pressure and the liquid pressure buildup in the
blanket during and after leachate injection events. The ﬁeld-scale study was not designed
to verify all parameters numerically evaluated by Khire and Haydar (2003) and Haydar

and Khire (2005). A complete veriﬁcation is beyond the scope of this study.

176

FIELD TEST SECTION AND METHODOLOGY
Field-scale testing of the GDL blanket used for leachate recirculation took place at an
MSW landﬁll located in the southern portion of Michigan. The landﬁll is currently active

and generates on an average of 45 m3 of leachate per day.

Test Section Layout

The GDL blanket was installed by placement of the GDL on the surface of an active cell
of the landﬁll. The GDL blanket was 34-m-long by 12-m-wide. A schematic of the GDL
blanket is presented in Figure 2. Before placing the GDL blanket, the surface of the
landﬁll cell was leveled. However, the surface could not be perfectly leveled. A
topographic survey of the blanket conducted after the placement of the GDL indicated
that the ground sloped towards east -— southeast and had an average slope of about 3.5%.
Below the surface, from the top to bottom, there is: (1) an average 50-mm-thick silty soil
(loess) used as a daily cover; (2) about 20-m-thick MSW; and (3) a leachate collection
and lining system. The GDL rolls used for this project were 4.6 m wide. Because the
blanket is about 12-m-wide, three GDL rolls were used with a 0.6-m overlap. In the
overlap zone, edges of the adjacent geonets were butted against each other and an overlap
of about 0.6 m was used for the top and bottom geotextiles.

The GDL used in this project had non-woven geotextile attached on one side and a
woven geotextile attached on the other side. Thus, we placed the GDL such that the non-
woven geotextile faced upward to prevent an intrusion of the waste into the geonet. The
woven geotextile side of the GDL was chosen to face downward to minimize potential
clogging due to the underlying silt layer. However, we believe that a GDL with non-

woven geotextiles on either side may be equally effective for the given application

177

 

1' " "' .—-' .5-

 

Leachate Storage

Location of one or more nested sensors with sensor ID (typical)

LEGEND:
E 4.5 m:

Control Valve

dance moisture content sensor (MCS)

impe
Piezometer with thermistor

M = Electrical

PIT

Pressure Gauge

TDR and impedance MCS with thermocouple

Leachate flow gauge
-->Direction of Surface Slope

TDR/MIT
FG

Vertical Pressure

Sensor

 

 

 

 

(a) 75 mm dia Perforated HDPE Pipe K

 

   
 
 

 

 

 

 

178

Not to Scale

  
 
 
  

 
 

 
      
  
 
     
 

   
     
          

    

    

    
    
         
            

           

   

   

   
 

   

 

'3293'3'3‘)
fang“? Mn
4 (3g ( 43434 I / x / / I5/3/5
iVVVVV»
5.3.3.3.».3.‘ \.\.\.~ ‘ . , . . .
(343(334343! . earner»?
43433634 1,4 r’e’r’ée’ee’z’r’
(363({31 ( . .W’M’I’I ”/3
(l(/( ’4 3 ‘3
3&6» ..... 2’
«At. . . .
$s$»$>;»»v ' ' '
Z3323?" / ‘ ‘ . '
’(‘t’t’e’z’z’a 4 I '
)VVVVV» '» ‘
)2’»)»»3 ‘o. :
3;»325» : t , r '
ﬁﬁ» V ,u-vr 5
23;». w; 323' ‘3
3 ‘ ' )5; ,3»: '3
4. aph»' 5
6". 23432522441. 4 a?"
P )VVVVVV»»%%%5
43' , , rye»)? >2 )5)»
r . ( £I29>3e>457325239573
4 . )4 ‘I/y) 3])?3}>)3}3)3
gyog‘gy wwwwmwwwh»
(3(‘(3( /r’-,.; > ’3' )5)» ‘ ' '
4 4 . ’1 . ri§2§i§>573237V
vrz/ ' /
, A.,syyssss‘yys
‘ p’sﬁsﬁﬁsﬂ‘s‘sfs

   

     

   
 

     
   
      
   

$555»;
7’23».

93252». //////
@3212 § '
)3: any)»;

   
  
 

I
2 2 2 9 23232325232323»
ss—s-wys-s-s-s-s-s
ﬁgﬁsgzgs‘sﬁﬁﬁfs‘s‘sﬁ
I e .

     

\~\~\~\~\o\.\VVVVVVVVV
\
e’

2 2»
'YVS’S‘S‘)“
( ' i '

  

 

but needs to be tested in the ﬁeld.

The key properties of the components of the GDL are presented in Table 1. A 12-
m-long perforated high density polyethylene (HDPE) leachate injection pipe having an
internal diameter equal to 75 mm was installed at the center of the blanket, parallel to the
width as per the schematic presented in Figure 2b. This pipe divided the test section into
two almost identical segments — east and west. The northern end of the perforated pipe
was capped and the southern end was connected to a hydraulic pump followed by a series
of three interconnected leachate storage tanks having a total storage capacity equal to
approximately 115 m3. A leachate ﬂow control valve, a digital pressure gauge, and a
magnetic ﬂow gauge were installed in the leachate pipe to control and monitor the

leachate head and ﬂow rate (Figure 2a).

Embedded Sensors

A total of 14 locations within the blanket were instrumented with these sensors:
impedance moisture content sensors; time domain reﬂectometry (TDR) moisture content
sensors; vibrating wire pressure transducers; and thermocouple and thermistor
temperature sensors. The key objective of these sensors was to monitor the migration of
injected leachate and the pressure and temperature changes in the blanket. The
installation locations and relevant details are presented in Figure 2.

The impedance moisture content sensors are designed to measure the electrical
impedance (R) between two electrodes embedded in a sand pack that is about 50 mm in
diameter (Gawande et al. 2003). The impedance of the sensor is inversely proportional to

the moisture content of the sand or the material surrounding the sand. A thermocouple of

179

 

 

 

Table 1. Physical Properties of the Geocomposite Drainage Material.

 

 

 

 

 

 

 

 

 

Upper ‘ Lower

component Geotextile Geotextile Geonet
Type Non-Woven Woven --
Thickness (mm) (ASTM D 5199) 2 0.5 5
Mass per Unit Area (g/mz) (ASTM D 5261) 270 200 --
Transmissivitya (cmz/s) (ASTM D 4716-00) -- -- 20
Permittivity(s'1) (ASTM D 4491) 1.5 1.1 --
Hydraulrc Conductrvrty (cm/s) (ASTM D 0.3 0.05 __
4491)
z115:;3513121)rent Opening Size (mm) (ASTM D 0.18 g 0. 6 ~ 125
Percent Open Area (%) (CW— 02215) -- ll 80

 

 

 

 

 

 

Note: a gradient of 0.1, normal load of 480 kPa, water (penneant) at 20 "C, between steel
plates for 15 minutes.

180

 

 

type T was added to the impedance moisture content sensor to allow the measurement of
temperature. The TDR moisture content sensor measures the surrounding medium’s
dielectric constant which is directly related to the moisture content. The TDR sensor used
in this study is about 685-mm—long and has a diameter equal to 19 mm.

The vibrating wire pressure transducer measures combined gas and liquid pressure.
The pressure transducer used in this study was not vented and required correction for
changes in the barometric pressure and temperature. A thermistor is attached to the
pressure transducer to allow measurement of the temperature and to correct the
measurement of the transducer. Unlike thermocouple sensors, thermistors measure the
absolute temperature and they do not require a reference temperature to make the
measurements. Haydar and Khire (2005c) have presented the details on the design,
function and application of the sensors used in this study. Haydar and Khire (2005c)
conducted lab experiments to determine the effect of electrical conductivity (BC) on the
measurement of the moisture content sensors. The TDR and impedance moisture content
sensors were backﬁlled with crushed glass (average particle diameter, D50 = 12 mm) and
submerged in saline solutions having EC ranging from 5 to 10 mS/cm. Potassium
Chloride was used an electrolyte to adjust the electrical conductivity of the solutions.
This range of EC represents the range of the EC of injected leachate in the ﬁeld. At
saturation, the TDR readings ranged from 1230 to 1300 uA and the impedance readings
ranged from 0.02 to 0.03 kQ.

The preﬁx of the location identiﬁcations presented in Figure 2a describes the
location of the sensor with respect to the leachate injection pipe (e.g., NW, W, etc.) and

the sufﬁx represents the perpendicular distance from the leachate injection pipe (e. g., 4.5

181

 

”E

..
‘ . “-.-

 

 

m, 12 m, etc.). At location B 4.5 m presented in Figure 2a, the sensor nest consists of the
following three types of sensors: (1) an impedance moisture content sensor with a
thermocouple; (2) a TDR moisture content sensor; and (3) a vibrating wire piezometer
with a thermistor. At location B 9 m, the sensor nest consists of an impedance moisture
content sensor and a TDR moisture content sensor. The use of different types of sensors
at the same location was to calibrate and independently verify the data measured among
sensors. A vertical pressure sensor was installed immediately outside the northern edge of
the test section to monitor the weight (or vertical stress) of waste placed on the test
section. The vertical pressure sensor also contained a thermistor which was used to
monitor the temperature of waste adjacent to the blanket. All sensors except for the
vertical pressure sensor were installed immediately below the GDL. Holes measuring 300
mm in diameter and about 300-mm-deep were excavated in the silty soil layer and waste
below the GDL and sensor(s) was placed in drainage backﬁll consisting of coarse sand or

crushed glass (D50 = 12 mm, hydraulic conductivity ~ 1 cm/s) as presented in Figure 2(b).

Monitoring System

Meteorological sensors including a rain gauge, an air temperature sensor, and a
barometric pressure sensor were also installed at the site. All sensors including the
leachate ﬂow gauge and pressure gauge were connected to a data logger located at the
site. The data logger was programmed to take readings at the desired frequency. Most of
the data collected from the test section was collected at a 5-minute frequency to allow for
a relatively precise monitoring of the injected leachate. From September 2003 to May
2004, about 1,800 m3 of leachate was recirculated in the GDL blanket corresponding to

approximately 27 leachate recirculation events. The leachate injection rate, Q, ranged

182

 

 

 

from 0.9 to 2.6 m3 per hour per meter length of the pipe perpendicular to the plane of the
paper (m3/hr/m). The control valve schematically presented in Figure 2a was used to
regulate the injection rate. The maximum leachate rate of 2.6 m3/hr/m corresponds to the
maximum rate the pump at the site could deliver at the total head that exists for the
system. During this monitoring period, the rain gauge recorded about 500 mm of

cumulative precipitation at the site.

RESULTS

Leachate Recirculation Trials

After the GDL blanket was covered with about 2-m-thick waste, leachate recirculation
was started in Sept. 2003. In Sept. 2003 the vertical stress recorded at the site was about

10 kPa. It rose to about 20 kPa in May 2004 due to waste ﬁlling.

Veriﬁcation of Response of Moisture Content Sensors

At location E 9 m, data collected from the TDR moisture content sensor was compared to
the impedance moisture content sensor. Data from this comparison is presented in Figure
3. In Figure 3, the negative values on the x-axis (-120 to 0) represent the time before the
leachate injection started and the positive values on the x-axis (0 to 240) represent the
time since the leachate injection began. The two arrows on the x-axis correspond to when
the leachate injection was started and turned off.

Figure 3 indicates that about 40 minutes after the leachate injection began, at
location B 9 m, the impedance started to drop and the TDR reading started to increase at

the same time indicating an arrival of the wetting front of the injected leachate. The

183

'o
._L

T""‘Y“_' -"Y"‘"“—Y"‘-Y“' 'T'TTT"T" ”T—"T "T"T"T‘T_T'T' r‘T‘T ""—7‘

Impedance Moisture Content
Sensor Reading at E 9 m,R (kn) o

.....................................................

Leachate Injection Duratio

_
'- I
. 1
_ jIJLIIlliI IIILJIIII‘J

 

 

"TTT‘ V‘ ‘V_' T‘ VT Th7? TIT—r—T‘TT'ﬂ

   

n} 0- 125 min

1 600

11400

51200

41000

on
o
o
TDR Moisture Content Sensor

 

.
i. . ‘ ‘
I I I i i L I 44

 

-120 -60 0 60 ‘1
Time (min)

Figure 3. Veriﬁcation of response of impedance moisture content sensor using TDR

moisture content sensor at location E 9 m.

184

20 180 240

Reading at E 9 m (M)

"I

impedance dropped from about 0.07 k!) to 0.03 kQ and stabilized. Similarly, the TDR

reading increased from about 500 to 1,200 DA and stabilized. According to the lab
calibration data, the stabilized high readings of the TDR and impedance moisture content
sensors correspond to 100% saturation immediately below the blanket. This indicates

temporary saturation of the blanket at Location E 9 m due to the leachate injection.

Verification of Response of Temperature and Pressure Transducer Sensors

Figure 4 presents the temperature and pressure head (hp) measurements at location E 4.5
m. Before leachate injection began, the temperature measured by the thermistor was
about 24°C and the pressure head measured by the transducer was about 10 cm. Leachate
was injected at a rate equal to 2.6 m3/hr/m. The net leachate injection pressure head (h,)
in the blanket was estimated to be around 4 m. It was estimated by subtracting head loss
in the segment of the leachate injection pipe between the blanket and the leachate
pressure gauge located outside the blanket. Note that the leachate injection pressure is
measured using a pressure gauge located approximately 40 m outside the blanket (Figure
2). The elevation of the leachate injection pressure measurement point is smaller than the
average elevation of the blanket by about 100 cm. Moody’s diagram (Moody 1944) was
used to estimate the head loss.

The temperature of the injected leachate was around 19°C. About 10 to 15 minutes
aﬁer the leachate injection began, in response to the arrival of the wetting front of the
injected leachate, the temperature measured by the thermistor decreased and

simultaneously the pressure head measured by the piezometer in the blanket increased.

185

 

97“ "3 it "-ﬂ

 

 

50,—. VI r'ﬁrY

T . I
” Net Injection Pressure ; ~ j
[ Head (h') 400 cm ~ ' _

40f;

 

19

 

: Leachate Pressure
Head (hp )

~-.«-. _,___,,.~-_ L..- .....s

(cm)

p
00
O
N
A

 

    

N
00
Temperature of Blanket at E 4.5 m (°C)

"VT"! j 'T" “I ' l "I T “ l

at E4.5 m, h
N
O

 

 

N
01

T S>nLeaChate Arrival

Injection Rate= 2. 6 m 3lhr/m
Leachate Injection Duration: 0- 100 min

0#_LDL Y .D...i ..... i27

-120 -60 60 120 180 240
Time (min)

 

 

Leachate Pressure Head in Blanket
._-_______.__.___I_

l
{—
l
r.
I»

 

Figure 4. Migration of leachate in the GDL blanket evaluated by change in temperature
and increase in pressure head measured by embedded sensor at location E 4.5 m.

186

Note that no precipitation was recorded during the time periods presented in
Figures 3 and 4. Hence, the response of the sensors observed in Figures 3 and 4 is strictly
due to the leachate injected in the blanket during those periods. Data presented in Figs. 3
and 4 indicate that moisture content, temperature, and pressure head, when measured
simultaneously, can be used to monitor the migration of injected leachate in or

immediately below the permeable blanket.

Flow of Leachate through the GDL Blanket

Figure 5 present the typical response of impedance moisture content sensors located in
the eastern and western portion of the GDL blanket for leachate injection events
corresponding to Q equal to 0.9 m3/hr/m. Figure 5 shows that the impedance moisture
content sensors that were initially partially saturated, experienced a decrease in
impedance as the injected leachate reached the location of the sensor. However, those
sensors that were initially completely saturated did not show any decrease in impedance.
Figure 5 represents the response of the impedance moisture content sensors to a leachate
injection event when about 20 m3 of leachate was injected in the blanket over a 125-
minute duration. Figure 5a shows that the injected leachate in the eastern portion of the
GDL reached the E 4.5 m sensor in 10 minutes, the SE 9 m sensor in 20 minutes, the E 9
m sensor in 50 minutes, the SE 16 m sensor in 60 minutes, and the NE 16 in sensor
within 120 minutes after the leachate injection began. Figure 5b shows that the injected
leachate in the western portion of the GDL reached both the NW 14 m and W 18 m in
100 minutes after the leachate injection began. The rate of travel of injected leachate in

the eastern and western portion of the blanket was not uniform. However, leachate did

187

 

0.01 - .--.,

L Eastern Portion of GDL Blanket
I Injection Rate = 0.9 m3/hr/m .
L Leachate Injection Duration: 0 - 125 min 5 E

 

 

 

 

a

r:

2 A

C C !

° :5

o 0‘ V +£4.5m
o

5 g; +E9m
H

.g :g +NE 9m
f g . - . +SE 9m
8 g L : : l 4 +514!“
'6 “ - : s

.3 5 «'M‘HA-"H'e’ \ Drop in . —>— NE 14m
n- U) ' V ' \ Impedance i

E ‘ ; +NE16m
- 01 . ___________ ..

+8516m

 

 

 

 

H3) 5 + 2 g 2 I
L“ '.;__1_L_J__.___‘ JLL‘Ligixglllii l I'I#_i

-120 -60 0 60 120 180 240
Time (min)

 

 

0.01"I*TTIYI'1jrrII.T.T-r,777,,,,-
Western Portion of GDL Blanket ‘

 

 

4 —9— NW 4.5m
i '—'<1— SW 4.5m

 

 

 

 

 

Impedance Moisture Content
Sensor Reading, R (kn)

 

3 ' 3 ‘3 . % Dropin +NW14m

0‘ I I ‘ Impedénce . —*—SW14m

1,4.--“ ' I _.._w.....
: Y . . ' Ti: T , j
. Injection Rate = 0.9 m3/hr/m j
(b) Leachate Injection Duration: 0 -125 min ,
111.1,1 11111 vrilllj.,,# LJLJL,1L4‘J_3

 

-120 -60 0 60 120 180 240
Time (min)

Figure 5. Response of impedance moisture content sensors located in the (a) eastern; and
(b) western portions of the blanket to a leachate injection event at leachate injection rate
equal to 0.9 m3/hr/m.

188

reach the farthest sensors (SE 16 m, NE 16 m, and W 18 m) located in the eastern and

western portions of the blanket.

Effect of Injection Rate on Leachate Travel

Figure 6 presents the arrival time of injected leachate at the locations of sensors in the
eastern and western portion of the GDL for two leachate injection events at Q equal to
0.9 and 2.6 m3/hr/m. The arrival time was determined by the response of impedance
moisture content sensors, located in the eastern and western portion of the GDL blanket,
to the arrival of injected leachate.

Moisture content sensors that were initially saturated were not able to record the
migration of injected leachate. Thus, only sensors that were partially saturated responded
to the migration of injected leachate. In order to ease the observation of the response of
sensors, sensors located in the same direction (e.g., NE, E, SE, etc.) were clustered and
assigned the same symbol in Figure 6. Hollow and solid symbols represent the arrival
times for leachate injection events at Q equal to 0.9 and 2.6 m3/hr/m, respectively. The
rate of travel of injected leachate was greater for the higher Q. Note that the slope of
dashed and solid lines in Figure 6 is the average rate of travel of injected leachate for Q
equal to 0.9 and 2.6 m3/hr/m, respectively. Figure 6 also shows that the solid symbols are
to the leﬁ side of alike hollow symbols, corresponding to the same cluster of sensors,
indicating an earlier arrival of injected leachate for the higher Q. For Q equal to 0.9
m3/hr/m, the injected leachate did not travel uniformly within the eastern or western
portions of the GDL blanket. For Q equal to 2.6 m3/hr/m, the travel of injected leachate
was more uniform. However, response of additional sensors is needed to conﬁrm such

conclusion.

189

 

. . GDL Test Section ; j
~ § Schematic ' North .

 

 

 

 

 

: Hollow Symbol 8 Dashed Line: Q- " 0. 9 m3/hr/m‘. j
Solid Symbol and Solid Line: Q= 2. 6 m3/hr/m
Oilrrrlil IllilllllilllLlIlIIlI
0 30 60 90 120 150

 

 

 

Distance of the Sensor from Injection Pipe (m)

Time Since Leachate Injection Started (minutes)

Figure 6. Effect of leachate injection rate on the rate of travel of leachate in the blanket

for leachate injection events corresponding to injection rates equal to 0. 9 and 2. 6
m 3/hr/m.

190

Based on the arrival time of injected leachate at a given sensor, the injected
leachate traveled at an average rate of 5 to 10 m/hr in the GDL blanket for Q equal to 0.9
m3/hr/m and at an average rate of 12 to 18 m/hr for Q equal to 2.6 m3/hr/m. We believe
that the rate of travel of injected leachate for a given leachate injection rate was not
uniform across the blanket due to these key factors: (1) preferential ﬂow in the GDL due
to wrinkles formed in the GDL during installation; (2) spatial variation of the hydraulic
conductivity and moisture content of the underlying waste; and (3) the waste surface
slightly sloped (~ 3.5%) towards east — southeast. Wrinkles were formed in the GDL
because the waste surface on which the GDL was laid was uneven and due to the thermal
stresses induced from a few days of sun exposure before the GDL was covered with

waste.

Correlation of Modeling and Field Results

Haydar and Khire (2005b) have presented a numerical parametric study of design
parameters inﬂuencing the rate of travel and pressure head of injected leachate in
permeable blankets. The conceptual model used in the parametric study consisted of an
LRS and a leachate collection system (LCS). The LRS consisted of a 60-m-wide and
0.15-m-deep PB made of pea gravel having a saturated hydraulic conductivity equal to
10'2 m/s. A 0.1-m-diameter perforated pipe was centrally placed in the blanket to inject
liquid. The LCS consisted of two leachate collection pipes horizontally spaced at 60 m
and embedded in a 0.3-m-deep drainage layer made of pea gravel.

Haydar and Khire (2005b) did not speciﬁcally simulate the GDL blanket presented
in this manuscript due to the limited scope of the study and limitations of the modeling.

For example, the GDL blanket having 10 mm thickness could not be accurately simulated

191

in HYDRUS-2D for the relatively large model domain (100-m wide by 22-m deep)
represented in the ﬁeld. The authors Showed that if the thickness of the PB is reduced, for
all other parameters kept constant, it results in an increase in the rate of travel of injected
liquid. Both, the simulated blanket and the GDL blanket had an equal hydraulic
transmissivity equal to 20 cmZ/s. Further details on the conceptual model and boundary
conditions are presented in Haydar and Khire (2005b). The rate of travel in the blanket
simulated by Haydar and Khire (2005b) (solid lines in Figure 7) and the measured rate of
travel from the ﬁeld data (hollow symbols in Figure 7) collected from the GDL blanket
are superimposed in Figure 7 for Q equal to 0.9 m3/hr/m. Because ﬁeld-scale hydraulic
conductivity of landﬁlled waste (k,,.) is rarely measured, a range of kw values (10'7 to 10'5
m/s) representing the typical range were Simulated (refer to the solid lines in Figure 7).

The initial degree of saturation of the blanket (S3) and the underlying waste (SW)
were assumed equal to 50% and 65%, respectively. In the ﬁeld, the degree of saturation
of the underlying waste was not measured. Haydar and Khire (2005b) Showed as the
degree of saturation of waste (or blanket) increases, the rate of travel of injected leachate
in the blanket also increases. However, this increase is relatively small. Hence, the
assumed values for the degrees of saturation, even though may not match the ﬁeld values,
are reasonable for the comparison sought here.

Figure 8 presents the simulated pressure head in the PB versus that measured in the
ﬁeld. The Simulated pressure head in the PB was measured at a distance of 4.5 m from
the injection pipe. This distance is equivalent to the distance between the embedded
pressure transducer and the injection pipe in the GDL blanket. For the simulation, Q was

assumed equal to 0.9 m3/hr/m and kw values ranging from 104’ to 10'5 m/S. S3 and SW

192

 

 

3O T—‘T‘r""‘rl"r r Tr -:iTj‘r“»—T 7,

 

 

 

 

 

 
   
  

i k :105 m/s
W

7; 3' Q = 0.9 m3/hr/m, K,3 :102 m/s

E -

:7 " Simulated Using HYDRUS-20 E] w j
d) l A NE v SE ,-
‘E 25 r 0 NW A

.2 O E

m -

.E 20 ‘~

0 7-

3 , 2 ‘

P 7 ‘7 k =10’7 m/s

iii 15 f 3 W .

E 3 ' j g,
3‘ a k :106 m/s
> L— - w _
a 10 _

L- .

I'-

o

H

II:

.c

o

«I

o

_I

a
O L___F__L.,_L__._.l__.l__. ;__i__I._._'_r-_i_ . l MLI r r l 1 1 l . i 1 1 1 l

0 30 60 90 120 150
Time Since Leachate Injection Started (minutes)

 

. Figure 7. Correlation between simulated and measured rate of travel of injected leachate
for leachate injection events corresponding to injection rates equal to 0.9 m3/hr/m.

193

 

1O 1" -~ -—--v----r— —— —--— ——-——- . _ -6 1
0'1 1 m3/hr/m& Slmulated-kw— 10 m/s ,

k8 - 3 x 10'2 m/s """ Simulated - kw = 10'5 mls

I Field Data - 26 Sep 2003
E1 Field Data - 09 Dec 2003

  

 

 

 

 

= 50% '

   

Pressure Head of Injected Leachate
in GDL Blanket, hp (m)

= 65% and S ____
0.01 1.. WW 1___ l. WWIWWI WW-
0 O. 5 1 1.5 2 2.5 3

Time since Leachate Injection Started (Hours)

Figure 8. Measured and Simulated pressure head of injected3 leachate 1n the GDL blanket
for two leachate injection events corresponding to Q= 0. 9 m 3/1'1r/m.

194

ranged from 50 to 90%. The magnitude of simulated h,p was less for a greater kw and for
lower SW and SB. From Sep. to Dec. 2003, the degree of saturation of the GDL blanket
and the waste beneath it increased due to the frequent leachate recirculation. This
correlated well with an increase in the pressure head measured in Dec 2003 compared to
that in Sep. 2003.

Because kw, SW, and SWCCS of the underlying waste are unknowns, the ﬁeld data
collected in this study cannot be simulated any more accurately. Figures 7 and 8 indicate
that the data measured in the ﬁeld is in the ball parks of simulated leachate travel for the
assumed variables. Hence, we believe that designers can use the modeling approach
presented by Haydar and Khire (2005b) to carry out preliminary design of ﬁeld-scale PBS
made up of GDLS or other permeable materials. However, we recommend
comprehensive monitoring of the PB to obtain site-speciﬁc data to improve the design

and operation of the PB.

Monitoring of Pressure Head and Temperature in the GDL Blanket

Potential clogging of the geotextiles in the GDL is one of the key operational concerns
for permeable blanket made of GDL. Haydar and Khire (2005b) numerically simulated
the clogging of the blanket by decreasing the hydraulic conductivity of the PB and
maintaining the leachate injection rate constant. The simulated data indicate that, if the
blanket partially clogs, the injection head needs to be increased to maintain the leachate
injection rate. As the leachate injection head increases, the liquid pressure heads in the
blanket and its immediate vicinity also increases. We evaluated potential clogging of the

GDL blanket during the monitoring period by evaluating the monitored leachate injection

195

pressures and leachate pressure head measured immediately below the GDL blanket at E
4.5 m.

Figure 9 presents the leachate injection rates and injection pressures during the

monitoring period. Figure 9 shows that the leachate injection pressures were about 200,
400, and 800 cm for leachate injection rates equal to 0.9, 1.4, and 2.6 m3/hr/m,
respectively, used at the site. For a given injection rate, leachate injection pressure
readings were constant and did not increase throughout the monitoring period. An
increase in these readings would have indicated potential clogging of the blanket.
Total hydraulic friction loss in the non-perforated 40-m-long portion of the leachate
injection pipe and the 12-m-long perforated portion of the leachate injection pipe was
estimated using Moody’s diagram (Moody 1944). For leachate injection rate ranging
from 0.9 to 2.6 m3/hr/m, the total friction loss ranged from about 100 to 300 cm. Thus,
the net leachate injection pressures (hi) ranged from about 50 to 400 cm.

For each leachate injection event, both the pressure transducer and the thermistor
located at E 4.5 m responded to the arrival of the injected leachate. Figure 10 presents the
pressure head and the temperature measured at E 4.5 m during the monitoring period. The
pressure transducer recorded an increase in leachate pressure head as the pore water
pressure increased due to the arrival of injected leachate. The increase in the pressure
head gradually dissipated after the leachate injection was stopped. The shorter peaks of hp
with a magnitude of about 30 cm in Figure 10 correspond to a leachate injection rate of
0.9 m3/hr/m and the taller peaks with a magnitude of about 80 cm correspond to a

leachate injection rate of 2.6 m3/hr/m.

196

ASE—5:: 3mm cozoﬁs 3203..

3 2

1 0

 

 

q a _I ﬁlJlllqulll

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0

l l
400 L ..
l
I
L

 

 

” I
m tIlIllIiIiItll... -II- .1--
.. n ..II.+ Iii - II.-
Od I .n H
.0“ a- u . .14
mm e .
. n
m 6 tab 8
. e r h n t
t U C C a
n a s a.AIVaR r
..n S e n
u m m L l -III1
” eP n
”IL H -
IrLIrLIL t-iL...t Ithn
0 0 0 O
0 O 0 0
8 2 6 O
1 1 2

E3 use: 9530...". nosed? Eugene...

3.65:3

Volga—21m _.

volcmwivr

moi>oziv_.

moiammtm _.

Figure 9. Leachate injection pressure and leachate injection rate measured at the site.

197

 

       
 

 

  
     

 

 

101 I ' i ‘ I 1 200 E
,— - - .1 Va
L ‘ .1:
~ - : . s"
15 L ................. ...... .................... 1 ................... J 150 IQ
A L F ,- V
g) . a ‘ II]
V Temperature at , . W - ‘5
3 X ‘ : (U
*5 20 """"""""" WasteTemperature’", 100 i,
3 5 Immediately Outside 0
a g , theBlanket 5
S k; hp at E 4.5 m ,1, 8

F ' , ‘- 'l!‘ , E
25 , ------------------ II ---------------- ,Lt‘ ------------ — 50 g

i" :I i '9 ‘- ,l I

. , , , .3
= a
30 I 1 - 0 0
m m x-r #- VI' .1

a a a s: a

o (B CD

<93 2. s 2. 2

:2 :2 z 52 "5

Figure 10. Leachate pressure heads and temperatures within the blanket and waste
temperature outside the blanket measured at the site.

198

Comparison of the leachate injection pressure and the leachate pressure head below the
blanket at E 4.5 m indicates that signiﬁcant leachate pressure head loss occurred as the
injected leachate traveled through the GDL blanket. The pressure head measured at E 4.5
m never exceeded the injection pressure head.

The temperature of the GDL blanket was also affected by leachate recirculation.
The temperature in the blanket temporarily decreased due to the arrival of injected
leachate. This occurred because the temperature of the injected leachate was almost
always less than the temperature of the blanket. The temperature of waste monitored
using the vertical stress sensor located immediately outside the northern edge of the
blanket was close to or greater than the temperature recorded at E 4.5 m (Figure 10). The
variation of the seasonal air temperature had a relatively small effect on the temperature

of the blanket due to the thermal insulation provided by the waste mass.

SUMMARY AND CONCLUSIONS
A 34-m-long by 12-m-wide permeable blanket constructed using a GDL was tested in the
ﬁeld to recirculate leachate in an active MSW landﬁll. Leachate was injected at rates
ranging from 0.9 to 2.6 m3/hr per meter length of the injection pipe. Moisture content,
temperature and pressure sensors were embedded immediately below the blanket to
monitor the rate of migration, temperature, and pressure of injected leachate in the
blanket.

Our major conclusions are: (1) the GDL blanket is a possible option to recirculate
leachate in MSW landﬁlls; (2) the rate ﬂow of injected leachate in the GDL blanket is a
function of the leachate injection rate, extent of wrinkles present in the GDL before it is

covered, the slope of the blanket, and the degree heterogeneity in the hydraulic properties

199

of underlying waste; (3) excess pressures were not developed in the vicinity of the
blanket indicating that the GDL blanket provided hydraulic continuity; (4) the moisture
content sensors embedded immediately below the blanket became saturated during
leachate injection demonstrating that the blanket transported the leachate; and (5) in the
long term, clogging of GDL blanket is a possibility, however, clogging at the site was not
indicated by the data during the 20-month-monitoring-period.

We recommend that designers consider the following when implementing this
design in the ﬁeld: (1) carry out numerical modeling of the blanket design using Site-
speciﬁc design variables; (2) evaluate the effect of increase in waste moisture content and
leachate pressure heads on the slope stability of the landﬁll; (3) instrument at least a
portion of the blanket in the ﬁeld to collect data on site-Speciﬁc performance; and (4)
maintain sufﬁcient distance between the edges of the blanket and the side Slopes of the
landﬁll to minimize the potential for leachate breakouts. Long-term monitoring of
pressure buildup and potential clogging still needs to be investigated to assess long-term

hydraulic efﬁciency of permeable blankets made up of GDL.

ACKNOWLEGEMENTS

This project was jointly funded by the Environmental Research & Education Foundation
(EREF) and Waste Management Inc. We sincerely appreciate the ﬁnancial support from
the Sponsors. Gundle/SLT Environmental, Inc. (GSE) donated the geocomposite drainage
material for this project. The authors would also like to express sincere appreciation to
Mr. Paul Mazanec, Mr. Ron Feldkamp, and Mr. Chester Stanley for their help during the
construction of the blanket and the coordination of numerous ﬁeld leachate recirculation

trials at the landﬁll Site. The results and opinions presented in this manuscript are those of

200

the authors. We also want to thank the three anonymous reviewers for their constructive

comments.

201

REFERENCES

Eith, A. and Koemer, R. (1992). “Field- evaluation of geonet ﬂow-rate (transmissivity)
under increasing load.” Geotextiles and Geomembranes, 11(4-6), 489-501.

Gawande, N., Reinhart, D., Thomas, P., McCreanor, P., and Townsend, T. (2003).
“Municipal solid waste in-situ moisture content measurement using an electrical
resistance sensor.” Waste Management, Vol. 23, 667-674.

Giroud, J., Zhao A., and Richardson, G. (2000). “Effect of thickness reduction on
geosynthetic hydraulic transmissivity.” Geosynthetics lntemational, 7(4-6), 433-
452.

Haydar, M.M. (2005). “Leachate recirculation in bioreactor landﬁlls: F ield-scale testing
and modeling.” Ph.D. Dissertation, Michigan State University, E. Lansing,
Michigan.

Haydar M.M. and Khire M.V. (2005a). “Leachate recirculation using horizontal trenches
in bioreactor landﬁlls.” Journal of Geotechnical & Geoenvironmental Engineering
Journal, ASCE, in press.

Haydar M.M. and Khire M.V. (2005b). “Leachate recirculation using permeable blankets
in bioreactor landﬁlls.” Journal of Geotechnical & Geoenvironmental Engineering
Journal, ASCE, in review.

Haydar M.M. and Khire M.V. (2005c). “Geotechnical sensor system to monitor injected
liquids in landﬁlls.” ASTM, Geotechnical Testing Journal, accepted.

Khire, M.V. and Haydar, M.M. (2003). “Numerical Evaluation of Granular Blankets for
Leachate Recirculation in MSW Landﬁlls.” Proceedings of the Ninth Sardinia Solid
Waste Conference, Cagliary, Italy, Oct.

Khire, M.V. and Haydar, M.M. (2005). “Leachate Recirculation using Geocomposite
Drainage Layer in Engineered MSW Landﬁlls.” Proceedings of GeoFrontiers’OS,
ASCE, Austin, TX, Jan.

Koemer, G.., Koemer, R., and Martin, J. (1994). “Geotextile Filters Used for Leachate

Collection Systems: Testing, Design of Field Behavior, Journal of Geotechnical
Engineering, ASCE, Vol. 120, No. 10, 1792-1803.

Koemer, R. (1999). Designing with Geosynthetics. Prentice Hall, New Jersey.
Koemer, R. (2000). “Leachate Recycling Loading to Bioreactor Landﬁlls for the Rapid

Degradation of Municipal Solid Waste.” Proceedings of ESD Solid Waste
Management Conference, MI, 39 pgs.

202

ml

McCreanor, P. (1998). “Landﬁll leachate recirculation systems: mathematical modeling
and validation.” Ph.D. Dissertation, University of Central Florida, Orlando, FL.

Moody, L. (1944). "Friction factor for pipe ﬂow." ASME Transactions. Vol. 66, 671.
Pohland, F. and Kim, J. (1999). “In situ aerobic treatment of Leachate in Landﬁll
Bioreactors.” Water Science & Technology, Vol. 40, 203-210.

Qian, X., Koemer, R., and Gray, D. (2002). Geotechnical Aspects of Landﬁll Design and
Construction. Prentice Hall, New Jersey.

Reinhart, D. and Al-Yousﬁ, B. (1996). “The impact of Leachate Recirculation on
Municipal Solid Waste Landﬁll Operating Characteristics.” Waste Management &
Research, Vol. 14, 337-346.

Reinhart, D. and Townsend, T. (1998). “Landﬁll Bioreactor Design & Operation”. Lewis
Publishers, Boca Raton, FL.

Simunek, J., Sejna, M., and Van Genuchten, M. (1999). “The HYDRUS-2D software
package for simulating the 2—D movement of water, heat, and multiple solutes in
variably Saturated media Version 2.0,” US. Salinity Laboratory, Agriculture
Research Service, USDA, Riverside, California.

203

SUMMARY AND PRACTICAL IMPLICATIONS

The key objectives of this dissertation were to: (1) perform a parametric study for
leachate recirculation or liquid injection using horizontal trenches (HTs) and permeable
blankets (PBS); and (2) evaluate the use of PBs for leachate recirculation in MSW
landﬁlls. To achieve these objectives, numerical modeling followed by ﬁeld-scale testing
of three instrumented PBS was carried out at an active MSW landﬁll located in Jackson,
Michigan.

The key installation related advantages of PBS over conventional systems are: (l)
excavation of waste is not needed during the construction of the blanket resulting in less
odor problems; and (2) a PB can substitute multiple horizontal trenches or vertical wells.
The key operational advantages of PBS are: (1) PB acts as an engineered heterogeneity
and reduces the effect of spatial variation of waste properties on wetting the waste
resulting in a relatively uniform distribution of injected leachate in the waste; and (2) PBs
made up of granular materials can provide an ideal platform to embed sensors for
monitoring the pressure, temperature, and migration of injected liquid to evaluate long-
term performance of the system.

The key findings of the numerical modeling of PBS and HTs and ﬁeld-scale

testing of PBS are presented below.

HORIZONTAL TRENCHES

The key ﬁndings of numerical modeling of HTS are:

0 An increase in the width of an HT results in a greater wetted width compared to when

the depth is increased.

204

0 When the hydraulic conductivity of an HT backﬁll is equal to or greater than that of
the surrounding waste, any further increase in the hydraulic conductivity of the HT
has negligible impact on the wetted area of waste.

0 For a given injection pressure head, the wetted width is independent of the saturated
hydraulic conductivity of the waste.

0 When two or more rows of HTS are designed, HTS in these rows should be vertically
staggered to reduce dry zones. HTS should be placed at about 10 m horizontal spacing

for formations consisting of HTs in a Single row for injection pressure head ranging

 

from 3 to 5 m. HTS should be placed at about 20 m horizontal spacing for formations
consisting of HTS in multiple rows where HTS are staggered and operated at injection
pressure head ranging from to 3 to 5 m.

0 The pattern of wetted area of injected leachate was greatly affected when
heterogeneity and anisotropy of the waste was considered. Introduction of
heterogeneity and anisotropy resulted in non-uniform leachate wetting. Increase in the
horizontal correlation length of hydraulic conductivity of waste resulted in a greater

wetting in the horizontal direction with a greater potential for leachate breakouts.

PERMEABLE BLANKETS
Numerical Modeling
The key ﬁndings from the numerical modeling of PBs are as follows.
0 Greater the leachate injection rate, hydraulic conductivity of the PB, or degree of

saturation of the waste and PB, faster the rate of travel of the injected leachate in

205

the PB. Similarly, lower the hydraulic conductivity of the underlying waste, faster
the rate of travel of the injected leachate in the PB

The pressure head of injected leachate in a PB is inversely proportional to the
hydraulic conductivities of the PB and waste. Hence, in order to maintain lower
pressure heads in the landﬁll resulting from the leachate injection, it is preferable
to select a material having relatively high hydraulic conductivity to construct the
PB.

Increase in the depth of a PB increases the storage capacity of the blanket and
hence decreases pressure head buildup of leachate in the blanket. Thus, a thicker
PB is preferable if slope stability of the landﬁll dictates lower leachate injection
pressure heads.

When leachate is injected periodically, the greater the ratio of on to off duration
of leachate injection, the greater the wetted width at steady-state. In addition,
when leachate is injected periodically, the average degree of saturation of the PB
and the underlying waste increases resulting in a faster travel of injected leachate

until a steady-state is reached.

Field-Scale Testing

The key ﬁndings from the ﬁeld-scale testing of the PBS are:

The migration of injected leachate in PBS can be monitored using an automated
sensing system designed to measure moisture content, temperature, and liquid

pressure within the PBS. However, unlike water content, pressure heads and

206

temperatures do not max out and hence can be used to detect the migration of
injected leachate in the PBS irrespective of the degree of saturation of the PBS.
The data collected from the sensors embedded in the three ﬁeld-scale permeable
blankets during the period from September 2003 to May 2005 indicated that the
injected leachate traveled across the entire width (~ 30 m) of the blankets. The
rate of travel of injected leachate was a function of leachate injection rate,
hydraulic conductivity or transmissivity of the PB, and degree of saturation of the
waste and PB. For a leachate injection rate of 3.6 m3/hr/m, injected leachate
traveled at an average rate of 18 m/hr in the glass blanket and at an average rate of
45 m/hr in the tires blankets. For a leachate injection rate equal to 2.6 m3/hr/m,
injected leachate traveled at an average rate of 15 m/hr in the GDL blanket.
Injected leachate traveled more uniformly in the glass PB compared to the other
two blankets. This was due to the uniform particle size distribution of crushed
glass used for the blanket. The migration of injected leachate in the GDL blanket
was not uniform possibly because of wrinkles in the GDL before it was covered
and Slight slope of the surface on which it was laid. In the shredded tires PB, the
ﬂow was not uniform due to the heterogeneous nature of the shredded tires.

The pressure head measured in the GDL blanket was greater than that measured
in the glass PB, at the same distance from the injection pipe. This was most likely
due to lower storage capacity in the GDL blanket compared to the glass PB.

The pore water pressure in none of the three PBS exceeded the leachate injection
pressure head. The maximum pressure head measured in the three PBS during the

monitoring period was 1.5 m.

207

 

 

0 The hydraulic efﬁciency of the blanket for long-term performance needs to be

investigated by further monitoring of the test sections.

208

 

IIIlIIj‘IjILIlILILIijIIlljjl