United States
Environmental Protection
Agency
Office of Solid Waste and
Emergency Response
Washington DC
EPA 540-R-04-007
OSWER 9283.1-26
April 2004
EPA (Draft) Technical Guidance
For RCRA/CERCLA Final Covers
Lateral Drainage Above CCL
Percolation Through CCL
1988 1989 1990 1991 1992 1993 1994 1995
by
Rudolph Bonaparte, Ph.D., P.E.
GeoSyntec Consultants
Atlanta, GA 30342
Beth A. Gross, P.E.
GeoSyntec Consultants
Austin, TX 78746
David E. Daniel, Ph.D., P.E.
University of Illinois
Urbana, IL 61801
Robert M. Koerner, Ph.D., P.E.
Drexel University/
Geosynthetic Research Institute
Philadelphia, PA 19104
Steve Dwyer, Ph.D., P.E.
Stephen F. Dwyer Engineering
Albuquerque, NM 87123
United States
Environmental Protection Agency
Office of Solid Waste and
Emergency Response
Washington DC
-------
Table of Contents
Chapter 1 Introduction 1-1
1.1 Overview 1-1
1.1.1 Purpose 1-1
1.1.2 Classification of Cover Systems 1-3
1.1.3 Organization of Document 1-7
1.2 Closure Regulatory Requirements 1-7
1.2.1 MSW Landfill Cover Systems 1-7
1.2.2 Hazardous Waste Landfill Cover Systems 1-10
1.2.3 Solid Waste Landfill Cover System Performance 1-11
1.2.4 CERCLA Site Cover Systems 1-12
1.2.5 Liquids Management Strategy 1-13
1.2.6 Design Life 1-15
1.2.7 Other Regulatory Requirements 1-15
1.3 Alternative Design Concepts and Materials 1-16
1.4 Gas Management Requirements 1-18
1.5 Typical Cover Systems Components 1-21
1.5.1 Surface Layer 1-21
1.5.2 Protection Layer 1-21
1.5.3 Drainage Layer 1-22
1.5.4 Hydraulic Barrier 1-23
1.5.5 Gas Collection Layer 1-23
1.5.6 Foundation Layer 1-24
1.6 Design Criteria Development and Conceptual Design 1-24
1.6.1 Overview 1-24
1.6.2 Regulatory Requirements 1-25
1.6.3 Climatic Criteria 1-25
1.6.4 Physical and Engineering Criteria 1-25
1.6.5 Aesthetic and Land Use Criteria 1-27
1.6.6 Ecological Criteria 1-28
Chapter 2 Individual Components of Cover Systems 2-1
2.1 Introduction 2-1
2.2 Surface Layer 2-1
2.2.1 General Issues 2-1
2.2.2 Elements of Design 2-2
2.2.2.1 Slope Inclination 2-2
2.2.2.2 Materials 2-4
2.2.2.2.1 Topsoil 2-4
2.2.2.2.2 Amended Topsoil 2-5
2.2.2.2.3 Soil-Gravel Mixture 2-6
2.2.2.2.4 Gravel Veneer 2-6
2.2.2.2.5 Riprap 2-7
2.2.2.2.6 Asphaltic Concrete 2-8
2.2.2.2.7 Other Materials 2-9
DRAFT - DO NOT CITE OR QUOTE
-------
2.2.2.3 Thickness 2-9
2.2.3 Vegetation 2-10
2.2.4 Surface-Water Control 2-13
2.2.5 Erosion Protection 2-18
2.2.5.1 Overview 2-18
2.2.5.2 Nature of Erosion 2-19
2.2.5.3 Short-Term and Long-Term Erosion 2-21
2.2.5.4 Sheet and Rill Erosion 2-23
2.2.5.4.1 Universal Soil Loss Equation 2-23
2.2.5.4.2 Water Erosion Prediction Project (WEPP) Model 2-24
2.2.5.5 Gully Erosion 2-25
2.2.5.5.1 Overview 2-25
2.2.5.5.2 Tractive Force Method for Vegetated Surface Layers 2-26
2.2.5.5.3 Horton/NRC Method for Vegetated Surface Layers 2-26
2.2.5.5.4 Permissible Velocity Method for Vegetated Surface Layers 2-27
2.2.5.5.5 Stephenson Method for Gravel or Riprap Surface Layers 2-28
2.2.5.6 Wnd Erosion 2-29
2.2.5.6.1 Revised Wnd Erosion Equation 2-29
2.2.5.6.2 Wind Erosion Prediction System 2-30
2.2.5.7 Erosion Control Materials 2-30
2.2.5.7.1 Temporary Erosion Control Materials 2-30
2.2.5.7.2 Permanent Erosion Control Materials 2-31
2.2.6 Construction 2-32
2.2.7 Maintenance 2-33
2.2.8 Monitoring 2-33
2.3 Protection Layer 2-33
2.3.1 General Issues 2-34
2.3.2 Elements of Design 2-34
2.3.2.1 Materials 2-34
2.3.2.2 Thickness 2-35
2.3.2.2.1 Desiccation Protection 2-36
2.3.2.2.2 Frost Penetration Protection 2-36
2.3.2.2.3 Accidential Human Intrusion Protection 2-38
2.3.2.2.4 Root Penetration Protection 2-39
2.3.2.2.5 Burrowing Animal Protection 2-40
2.3.2.2.6 Vegetation Support 2-41
2.3.2.2.7 Water Storage 2-41
2.3.2.2.8 Radon Attenuation 2-44
2.3.3 Construction 2-44
2.3.4 Maintenance 2-45
2.3.5 Monitoring 2-45
2.4 Drainage Layer 2-45
2.4.1 General Issues 2-45
2.4.2 Elements of Design 2-46
2.4.2.1 Materials 2-46
2.4.2.1.1 Granular Materials 2-46
2.4.2.1.2 Geosynthetics 2-48
DRAFT - DO NOT CITE OR QUOTE
-------
2.4.2.2 Thickness of Granular Layers 2-50
2.4.2.3 Required Flow Capacity 2-50
2.4.2.4 Drainage Layer Outlets 2-52
2.4.3 Construction 2-52
2.4.4 Maintenance 2-52
2.4.5 Monitoring 2-52
2.5 Hydraulic Barrier 2-53
2.5.1 General Issues 2-53
2.5.2 Elements of Design 2-53
2.5.2.1 Materials 2-54
2.5.2.1.1 GMs 2-54
2.5.2.1.2 GCLs 2-55
2.5.2.1.3 CCLs 2-59
2.5.2.2 Thickness 2-60
2.5.2.2.1 GMs 2-60
2.5.2.1.2 GCLs 2-60
2.5.2.1.3 CCLs 2-61
2.5.2.3 Percolation 2-61
2.5.2.4 Gas Containment 2-63
2.5.2.5 Differential Settlement 2-63
2.5.2.6 Wet-Dry Cycles 2-65
2.5.2.7 Freeze-Thaw Cycles 2-66
2.5.2.8 Shear Strength 2-67
2.5.2.9 Accidential or Intentional Puncture 2-70
2.5.2.10 Anticipated Lifetime 2-70
2.5.2.10.1 GMs 2-70
2.5.2.10.2 GCLs 2-72
2.5.2.10.3 CCLs 2-73
2.5.3 Composite Hydraulic Barriers 2-73
2.5.3.1 Prompt Placement of Overlying Materials 2-74
2.5.3.2 Intimate Contact 2-74
2.5.4 Construction 2-75
2.5.5 Maintenance 2-76
2.5.6 Monitoring 2-77
2.6 Gas Collection Layer 2-77
2.6.1 General Issues 2-77
2.6.2 Elements of Design 2-78
2.6.2.1 Materials 2-78
2.6.2.1.1 Granular Materials 2-78
2.6.2.1.2 Geosynthetics 2-79
2.6.2.2 Thickness of Granular Layers 2-80
2.6.2.3 Required Flow Capacity 2-80
2.6.3 Gas Collection Layer Outlets 2-80
2.6.4 Construction 2-81
2.6.5 Maintenance 2-81
2.6.6 Monitoring 2-81
2.7 Foundation Layer 2-81
2.7.1 General Issues 2-81
2.7.2 Elements of Design 2-82
2.7.2.1 Materials 2-82
2.7.2.2 Thickness 2-83
DRAFT - DO NOT CITE OR QUOTE
-------
2.7.3 Construction 2-83
2.7.4 Maintenance 2-83
2.7.5 Monitoring 2-83
2.8 Examples of Cover Systems for Different Applications 2-83
Chapter 3 Alternative Design Concepts and Materials 3-1
3.1 Introduction 3-1
3.2 ET Barrier Design 3-3
3.2.1 Overview 3-3
3.2.2 General Issues 3-6
3.2.3 Elements of Design 3-6
3.2.4 Design Concept 3-6
3.2.5 Soil Thickness 3-7
3.3 Capillary Barrier Design 3-8
3.3.1 Overview 3-8
3.3.2 General Issues 3-10
3.3.3 Elements of Design 3-10
3.3.4 Design Concept 3-11
3.3.5 Coarser-Grained Soil Layer 3-13
3.3.6 Internal Stability 3-14
3.4 Alternate Design Performance Evaluation 3-14
3.4.1 Introduction 3-14
3.4.2 Numerical Modeling 3-15
3.4.3 Performance Monitoring 3-16
3.4.4 Natural Analogs 3-17
3.5 Construction of Alternative Designs 3-20
3.5.1 Compaction Requirements 3-20
3.5.2 Capillary Barrier Soil Interfaces 3-20
3.6 Maintenance and Monitoring of Alternative Designs 3-21
3.6.1 Maintenance 3-21
3.6.2 Monitoring 3-21
3.7 Alternative Materials 3-22
3.7.1 Geofoam 3-22
3.7.2 Shredded Tires 3-22
3.7.3 Sprayed Elastomers 3-25
3.7.4 Paper Mill Sludges 3-26
Chapter 4 Hydraulic Analysis and Design 4-1
4.2 Introduction 4-1
4.2 Characteristics of Water Balance Models 4-1
4.2.1 Overview 4-1
4.2.2 Water Balance Concept 4-1
4.2.3 Water Balance Methods 4-3
4.2.3.1 Simplified Manual Method 4-4
4.2.3.2 HELP 4-10
4.2.3.3 LEACHM Model 4-13
4.2.3.4 UNSAT-H 4-14
4.2.3.5 SoilCover 4-15
DRAFT - DO NOT CITE OR QUOTE jv
-------
4.2.3.6 HYDRUS-2D 4-18
4.3 Evaluation of Water Balance Models 4-19
4.3.1 Overview 4-19
4.3.2 Lysimeters at DOE Hanford Site 4-19
4.3.3 Test Plots at Hill Air Force Base 4-23
4.3.4 Test Plots in Live Oak, Georgia and Wenatchee, Washington 4-27
4.4 Recommendations for Application of Water Balance Models 4-32
4.5 Design of Drainage Layers 4-34
4.5.1 Simplified Manual Method 4-34
4.5.2 Refinement to Simplified Manual Method 4-37
4.5.3 HELP Model 4-38
4.6 Design of Slope Transitions 4-39
4.7 Design of Filter Layers 4-42
4.7.1 Overview 4-42
4.7.2 Soil Filters 4-42
4.7.3 GT Filters 4-42
Chapters Gas Emission Analysis and Collection System Design.5-1
5.1 Introduction 5-1
5.2 Mechanisms of Gas Generation and Emission 5-1
5.2.1 Overview 5-1
5.2.2 MSW Landfill Gas Generation 5-3
5.2.3 Gas Emissions 5-7
5.3 Characteristics of Selected Gas Emission Models 5-10
5.3.1 LandGEM Model for MSW 5-10
5.3.2 Diffusion Model for Emissions of Organic Vapors 5-12
5.4 Design of Gas Collection Systems 5-12
Chapter 6 Geotechnical Analysis and Design 6-1
6.1 Introduction 6-1
6.2 Static Slope Stability 6-1
6.2.1 Overview 6-1
6.2.2 Limit Equilibrium Analyses 6-3
6.2.2.1 Overview 6-3
6.2.2.2 Infinite Slope 6-4
6.2.2.3 Slope of Finite Length 6-7
6.2.3 Stress-Deformation Analyses 6-12
6.2.4 Shear Strength Parameters 6-13
6.2.5 Construction Considerations 6-19
6.2.6 Factors of Safety 6-19
6.3 Seismic Slope Stability and Deformation 6-25
6.3.1 Overview 6-25
6.3.2 Seismic Hazard Evaluation 6-26
6.3.3 Seismic Response Analysis 6-30
6.3.3.1 Introduction 6-30
6.3.3.2 Material Properties Selection 6-30
6.3.3.3 Simplified Response Analysis 6-33
DRAFT - DO NOT CITE OR QUOTE
-------
6.3.3.4 Analytical and Numerical Response Analyses 6-35
6.3.4 Dynamic Shear Strength 6-37
6.3.5 Seismic Stability and Deformation Analysis 6-37
6.3.5.1 Overview 6-37
6.3.5.2 Psuedo-Static Factor of Safety Method 6-37
6.3.5.3 Modified Pseudo-Static Factor of Safety Method 6-38
6.3.5.4 Permanent Seismic Deformation Method 6-40
6.3.5.5 Seismic Deformation Performance Criteria 6-43
6.4 Settlement 6-45
6.4.1 Mechanisms of Settlement 6-45
6.4.2 Settlement of Foundation Soils 6-45
6.4.3 Overall Waste Compression 6-46
6.4.4 Differential Settlement Due to Localized Mechanisms 6-49
6.4.5 Impacts of Settlement on Cover Systems 6-50
6.5 Steep Slopes 6-51
6.5.1 Introduction 6-51
6.5.2 Waste Buttress 6-51
6.5.3 Steep Cover System Slopes 6-58
6.6 Soft Waste Materials 6-60
Chapter 7 Lessons Learned 7-1
7.1 Introduction 7-1
7.2 Soil Barriers 7-2
7.2.1 Test Plots in Omega Hills, Wisconsin 7-2
7.2.2 Test Plots in Kettleman City, California 7-5
7.2.3 Cover Systems in Maine 7-6
7.2.4 Test Plots in Live Oak, Georgia and Wenatchee, Washington 7-8
7.2.5 Test Plots in Hamburg, Germany 7-8
7.2.6 Test Plots in Albuquerque, New Mexico 7-12
7.2.7 Test Plots in Los Alamos, New Mexico 7-15
7.3 GM Barriers 7-16
7.3.1 Percolation through GM Barriers 7-16
7.3.2 GM Barrier Seam Problem Due to Contamination 7-17
7.3.3 GM Barrier Seam Problem Due to Moisture 7-18
7.3.4 Temperature Fluctuations During GM Installation 7-18
7.3.5 Fate of GM Wrinkles 7-20
7.4 Slope Stability 7-22
7.4.1 Overview 7-22
7.4.2 Cover System Slope Failure During Construction 7-22
7.4.3 Cover System Slope Failure After Rainfall or Thaw 7-23
7.4.4 Soil Cover Damage Due to Earthquakes 7-29
7.4.5 Results of EPA GCL Test Plots 7-30
7.5 Waste Settlement 7-37
7.6 Stormwater Management and Erosion Control 7-39
7.6.1 Failure of Erosion-Mat Lined Downchute 7-39
7.6.2 Excessive Erosion and Gullying 7-40
7.6.3 Failure of Surface-Water Runoff Collector 7-42
7.7 Gas Pressures 7-44
DRAFT - DO NOT CITE OR QUOTE vi
-------
7.8 Miscellaneous Problems 7-46
Chapter 8 Performance Monitoring 8-1
8.1 Introduction 8-1
8.2 Infiltration Monitoring 8-3
8.2.1 Overview 8-3
8.2.2 Leachate Collection System Monitoring 8-3
8.2.3 Drainage Layer Monitoring 8-4
8.2.4 Lysimeter Monitoring 8-5
8.3 Soil Moisture Monitoring 8-7
8.3.1 Overview 8-7
8.3.2 Neutron Probes 8-8
8.3.3 Time Domain Reflectometry 8-10
8.3.4 Frequency Domain Reflectometry 8-12
8.3.5 Tensiometers 8-13
8.3.6 Electrical Resistance Sensors 8-14
8.3.7 Thermocouple Psychrometers 8-16
8.3.8 Heat Dissipation Sensors 8-16
8.4 Gas Emissions Monitoring 8-17
8.5 Settlement Monitoring 8-19
Chapter 9 Maintenance and Long-Term Issues 9-1
9.1 Introduction 9-1
9.2 Cover System Maintenance 9-8
9.2.1 Overview 9-8
9.2.2 Vegetation-Related Maintenance 9-9
9.2.3 Erosion-Related Maintenance 9-9
9.2.4 Subsidence-Related Maintenance 9-9
9.2.5 Other Surface Layer Related Maintenance 9-10
9.2.6 Drainage Layer Related Maintenance 9-10
9.2.7 Maintenance of Surface-Water Management System 9-10
9.2.8 Maintenance of Cover Monitoring System 9-11
9.3 Site End Use 9-11
9.3.1 Overview 9-11
9.3.2 Ecological Reuse 9-12
9.3.3 Recreational Reuse 9-12
9.3.4 Industrial and Commercial Reuse 9-13
9.3.5 Case Histories 9-15
Appendix A: References
DRAFT - DO NOT CITE OR QUOTE vi i
-------
List of Acronyms
ACAP
ALCD
ACS
ARAR
ASTM
BOD
BTEX
C&DW
CAA
CCL
CERCLA
CFR
CH
CL
COD
CQA
CQC
CSPE
CSPE-R
DOE
EG
EIA-R
EPA
EPS
ET
FDEP
FDR
FHWA
FID
FOS
FS
fPP
fPP-R
GC
GCL
GM
GN
GPR
GPS
GT
HAP
HOPE
Alternative Cover Assessment Program
Alternative Landfill Cover Demonstration
apparent opening size
applicable or relevant and appropriate requirement
American Society for Testing and Materials
biological oxygen demand
benzene, toluene, ethylbenzene, and xylenes
construction and demolition waste
Clean Air Act
compacted clay liner
Comprehensive Environmental Response, Compensation and
Liability Act
U.S. Code of Federal Regulations
high-plasticity clay (according to USCS)
low-plasticity clay (according to USCS)
chemical oxygen demand
construction quality assurance
construction quality control
chlorosulfonated polyethylene
chlorosulfonated polyethylene - reinforced
U.S. Department of Energy
emissions guidelines
ethylene interpolymer alloy-reinforced
U.S. Environmental Protection Agency
expanded polystyrene
Evapotranspiration
Florida Department of Environmental Protection
frequency domain reflectometry
Federal Highway Administration
flame ionization detector
filtration opening size
factor of safety
flexible polypropylene
flexible polypropylene reinforced
geocomposite
geosynthetic clay liner
geomembrane
geonet
ground penetrating radar
global positioning system
geotextile
hazardous air pollutants
high density polyethylene
DRAFT - DO NOT CITE OR QUOTE
VIM
-------
HELP = Hydrologic Evaluation of Landfill Performance
HW = hazardous waste
HSWA = Hazardous and Solid Waste Amendment
ISM = instantaneous surface monitoring
ISS = integrated surface sampling
IW = industrial waste
L = Liters
LandGEM = EPA Landfill Gas Generation Model
LCRS = leachate collection and removal system
LDLPE = low density linear polyethylene
LDR = Land Disposal Restrictions
LDS = leak detection system
LE = limit equilibrium
LEACHM = Leachate Estimation and Chemistry Model
LLDPE = linear low density polyethylene
LMDPE = linear medium density polyethylene
Iphd = liter/hectare/day (1 Iphd = 9.35 gallon/acre/day (gpad))
MACT = maximum achievable control technology
MCL = maximum contaminant level
MSE = mechanically stabilized earth
MSW = municipal solid waste
MSWLF = municipal solid waste landfill
NCDC = National Climatic Data Center
NCP = National Contingency Plan
NESHAP = National Emission Standards for Hazardous Air Pollutants
NMOC = non-methane organic compound
NPDES = National Pollution Discharge Elimination System
NRC = U.S. Nuclear Regulatory Commission
NRCS = National Resources Conservation Service
NSPS = New Source Performance Standards
Oil = Operating Industries, Inc.
OU = operable unit
PCB = polychlorinated biphenyl
PCDD = polychlorinated dibenzo-p-dioxins
PCDF = polychlorinated dibenzo-furans
PE = polyethylene
PERM = permanent erosion and revegetation material
PET = potential evapotranspiration
PHGA = peak horizontal ground acceleration
PMP = probable maximum precipitation
PPL = priority pollutant list
ppm = parts per million
PVC = polyvinyl chloride
QA = quality assurance
QC = quality control
ROD = Record of Decision
DRAFT - DO NOT CITE OR QUOTE
IX
-------
RCRA = Resource Conservation and Recovery Act
RCPS = rigid cellular polystyrene
RUSLE = Revised Soil Loss Equation
RWEQ = Revised Wind Erosion Equation
SARA = Superfund Amendments and Reauthorization Act
SASW = spectral analysis of surface waves
SC = clayey sand (according to USCS)
SCS = USDA Soil Conservation Service
SDRI = sealed double-ring infiltrometer
SE = southeast
SITE = Superfund Innovative Technology Evaluation
SMCL = secondary maximum contaminant level
SVOC = semivolatile organic compound
SVT = solvent vapor transmission
SWRRB = Simulation for Water Resources in Rural Basins
TDR = time domain reflectometry
TDS = total dissolved solids
TERM = temporary erosion and revegetation material
TR-55 = Technical Release 55 (SCS, 1986)
TRM = turf reinforcement mat
TSS = total suspended solids
TOC = total organic carbon
TOC = total organic compound (in Chapter 5)
UMTRA = Uranium Mill Tailings Remedial Action
UMTRCA = Uranium Mill Tailings Radiation Control Act
USCS = Unified Soil Classification System
USDA = United States Department of Agriculture
USFWS = U.S. Fish and Wildlife Service
USGS = U.S. Geological Survey
USLE = Universal Soil Loss Equation
VFPE = very flexible polyethylene
VLDPE = very low density polyethylene
VOA = volatile organic acid
VOC = volatile organic compound
WES = U.S. Army Corps of Engineers Waterways Experiment Station
WVT = water vapor transmission
XPS = extruded polystyrene
DRAFT - DO NOT CITE OR QUOTE
-------
List of Variables
A = dimensionless parameter (dimensionless)
A = area of emitting source (m2) (in Chapter 5)
Ab = area of drainage basin or subbasin per basin or subbasin width
(m2/m)
As = average annual soil loss by sheet and rill erosion (tonnes/ha/yr)
aa = cohesion (for internal strength) or adhesion (for an interface) for the
critical potential slip surface above the hydraulic barrier (Pa)
aa; = apparent adhesion (for an interface) or cohesion (for internal
strength) for the critical potential slip surface (Pa), as defined in
Figure 6-8
ab = cohesion (for internal strength) or adhesion (for an interface) for the
critical potential slip surface below the hydraulic barrier (Pa)
a; = adhesion (for an interface) or cohesion (for internal strength) for the
critical potential slip surface (Pa)
B = dimensionless parameter (dimensionless)
B = Distance over which differential settlement, A, occurs (m)
C = vegetative cover and management factor (dimensionless)
Cd = empirical factor (dimensionless)
Ce = void ratio correction factor (dimensionless)
CF = vegetal cover factor (dimensionless)
Ci = vegetal retardance curve index (dimensionless)
Cr = runoff coefficient (dimensionless)
Cs = surface layer coefficient (dimensionless)
CN = runoff curve number (dimensionless)
COG = combined crop factors (dimensionless)
Cae = modified secondary compression index (dimensionless)
Caei = modified secondary compression index during the intermediate
secondary compression period (dimensionless)
Ca£2 = modified secondary compression index during the long-term
secondary compression period (dimensionless)
Ci,i-Ci,2 = concentration gradient of species i (Mg/m3)
c = runoff coefficient (dimensionless)
cs = cohesion of soil material above the critical potential slip surface (Pa)
D = flow depth (m)
D; = depth of influence (m)
D; = diffusivity of species i through cover material (m/yr2) (in Chapter 5)
DIS = particle size at which 15% by dry weight of the soil particles are
smaller (mm)
D50 = minimum gravel or riprap mean particle size to withstand the peak
rate of runoff (mm)
D85 = particle size at which 85% by dry weight of the soil particles are
smaller (mm)
DRAFT - DO NOT CITE OR QUOTE xi
-------
d = depth of rainfall in time of concentration from a storm with a certain
return period (m)
E = equivalency factor (dimensionless)
Ev = vertical evaporative flux (mm/day)
EF = erodible fraction (dimensionless)
ER; = mass emission rate of species i (Mg/yr)
ET = evapotranspiration (mm/day)
F = flow concentration factor (dimensionless)
Fw = seepage force (N)
FS = factor of safety (dimensionless)
FSA = factor of safety for critical potential slip surface above the hydraulic
barrier (dimensionless)
FSe = factor of safety for critical potential slip surface below the hydraulic
barrier (dimensionless)
FSmin = minimum acceptable factor of safety (dimensionless)
f(S) = slope function (dimensionless)
fw = seepage force per unit volume (N/m3)
G = dynamic shear modulus (Pa)
G/Gmax = dynamic shear modulus reduction factor (dimensionless)
Gs = specific gravity of gravel or riprap (dimensionless)
Gmax = maximum small-strain dynamic shear modulus (Pa)
g = acceleration of gravity (m/s2)
H = height of the falling weight (m)
Hf = elevation difference along flow path (m)
Hs = soil layer thickness (m)
Hw = depth of water that can be stored in a soil layer for subsequent
removal by plants
HI = height of waste at time ti (m)
H2 = height of waste at time t2 (m)
AHS = secondary waste settlement (m)
h = height of slope (m), as defined in Figure 6-4
ha = relative humidity of the air (dimensionless)
havg = average hydraulic head (m)
hm = maximum head in drainage layer (m)
hu = height of slope above the slope grade break (m), as illustrated in
Figure 6-6
hr = relative humidity at the soil surface (dimensionless)
hz = minimum head at which flow into the coarser-grained layer first
occurs (m)
I = infiltration into surface cover soil (mm/day)
i = hydraulic gradient (dimensionless)
ir = rainfall intensity (m/s)
K = soil erodility factor (dimensionless)
K' = soil roughness factor (dimensionless)
k = hydraulic conductivity (m/s)
DRAFT - DO NOT CITE OR QUOTE xi i
-------
k = methane generation rate constant (yr"1) (in Chapter 5)
kcs = cover soil saturated hydraulic conductivity (m/s)
kd = drainage layer hydraulic conductivity (m/s)
kds = granular drainage layer hydraulic conductivity (m/s)
kf = long-term field hydraulic conductivity of granular drainage layer
(m/s)
kg = gas conductivity (m/s)
kh = pseudo-static seismic coefficient (dimensionless)
ki = laboratory hydraulic conductivity of granular drainage layer (m/s)
kn = cross-plane hydraulic conductivity of geotextile (m/s)
ks = saturated hydraulic conductivity (m/s)
ku = unsaturated hydraulic conductivity (m/s)
ky = pseudo-static seismic coefficient that produces a psuedo-static slope
stability FS of 1.0 (dimensionless)
kyg = yield acceleration (m/s2)
L = lateral drainage (mm/day)
Ld = length of drainage layer flow path (m)
Lf = length of overland flow path (m)
Lfg = thickness of finer-grained soil layer (m)
L0 = methane generation potential (m3/Mg)
LS = slope length and steepness factor (dimensionless)
/ = slope length (m)
M; = mass of solid waste in the ith section (Mg)
MW = earthquake moment magnitude (dimensionless)
n = Manning's roughness coefficient for the considered vegetative cover
(dimensionless)
np = porosity of gravel or riprap layer (dimensionless)
ns = Manning's roughness coefficient for the bare soil (dimensionless)
095 = the 95% opening size of the geotextile (mm)
Pa = vapor pressure in the air above the evaporating surface (Pa)
Pc = conservation support practice factor (dimensionless)
P = precipitation (mm/day)
PERC = percolation through the cover system (mm/day)
PERC* = percolation through the cover soil (mm/day)
PET = potential evapotranspiration (mm/day)
Q(x) = mass transport of soil at downwind distance x (kg/m)
QM = maximum expected gas generation flow rate (Mg/yr)
Qmax = mass transport of soil (kg/m)
Q(x)max = maximum mass transport of soil at downwind distance x (kg/m)
Q = peak rate of runoff (m3/s/m)
qc = flow capacity of drainage layer (m3/s/m)
qm = maximum flow rate in drainage layer (m3/s/m)
R = runoff (mm/day)
Re = rainfall energy/erosivity factor (dimensionless)
Rf = permissible velocity reduction factor (dimensionless)
Rn = net radiant energy available at the surface (mm/day)
DRAFT - DO NOT CITE OR QUOTE x\\\
-------
S = slope inclination (dimensionless)
S(z,t) = Sink term representing uptake by transpiration (s"1)
SCF = soil crust factor (dimensionless)
Sr = retention parameter (mm/day)
s(x) = field length scale (m)
T = geosynthetic tension above the potential slip surface (N/m)
t = thickness of material above the critical potential slip surface (m) (in
Chapter 6)
t = time (s) (in Chapter 4)
ta = thickness of soil layer at point A (m), as defined in Figure 6-5
tavg = average thickness of soil layer between points A and B, which are
defined in Figure 6-5 (m)
tb = thickness of soil layer at point B (m), as defined in Figure 6-5
tc = time of concentration (s)
td = drainage layer thickness (m)
tds = granular drainage layer thickness (m)
t; = age of the ith section (yr)
tm = required thickness of the internal drainage layer (m)
tw = thickness of water flow parallel to the slope (m), as defined in Figure
6-3
t w = thickness of water in Wedge 1 (m), as defined in Figure 6-4;
ti = starting time for the period of secondary compression (s)
t2 = ti plus the time duration of secondary compression or intermediate
secondary compression (s)
t3 = T2 plus the time duration of long-term secondary compression (s)
Ua = wind speed (km/hr)
v = flow velocity (m/s)
vs = shear wave velocity of material (m/s)
vs, waste = shear wave velocity of waste (m/s)
W = mass of the falling weight (tonne)
Wb = buoyant unit weight (N)
WF = weather factor (kg/m)
x = downwind distance (m)
x = cover thickness (m) (in Chapter 5)
xc = critical distance along a slope before gully formation begins (m)
z = vertical coordinate (m)
F = slope of the saturation vapor pressure versus temperature curve at the
mean temperature of the air (dimensionless)
*F = geotextile permittivity (s-1)
\j/ = matric potential (negative) due to capillary suction forces (N/m2)
a = empirical constant (m/tonne)0'5
(3 = slope angle (degrees)
yb = average buoyant unit weight of material above the critical potential
slip surface (N/m3)
ysat = average saturated unit weight of material above the critical potential
slip surface (N/m3)
DRAFT - DO NOT CITE OR QUOTE XJV
-------
Yt
Yt, waste
Yw
A
AWfoliage
AWsoli
AWsurface
5
e
ea
Qafc
fallow
eds
6fc
eh
QSC
Quit
Uw
Pa
Pw
On
T
Ta
tab
Tab
Te
V
total unit weight of material above the critical potential slip surface
or total unit weight of material (N/m3)
total unit weight of waste (N/m3)
unit weight of water (N/m3)
differential settlement (m)
change in water storage on plant foliage (mm/day)
change in water storage in cover system soil (mm/day)
change in water storage at surface (mm/day)
shear displacement (m)
soil volumetric moisture content (dimensionless)
air transmissivity (m3/s/m)
soil apparent field capacity (dimensionless)
allowable hydraulic transmissivity of geosynthetic drainage layer
(m3/s/m)
geosynthetic drainage layer transmissivity (m3/s/m)
granular drainage layer transmissivity (m3/s/m)
soil field capacity (dimensionless)
hydraulic transmissivity (mVs/m)
soil water storage capacity (dimensionless)
ultimate hydraulic transmissivity of geosynthetic drainage layer
(m3/s/m)
soil wilting point (dimensionless)
pore size distribution index (dimensionless)
air viscosity (kg/m/s)
water viscosity (kg/m/s)
air density (kg/m3)
water density (kg/m3)
normal stress (kPa)
shear stress (Pa)
allowable shear stress (kPa)
allowable shear stress for the surface layer with bare soil (kPa)
allowable shear stress for the Horton/NRC method (kPa)
effective shear stress applied to the surface layer by the flowing
water (kPa)
psychrometric constant (dimensionless)
angle of repose or gravel or riprap (degrees)
angle of internal or interface friction for the critical potential slip
surface (degrees)
angle of internal or interface friction for the critical potential slip
surface above the hydraulic barrier (degrees)
angle of internal or interface friction for the critical potential slip
surface below the hydraulic barrier (degrees)
angle of internal friction for the soil material (i.e., protection layer
and/or granular drainage layer) above the critical potential slip
surface (degrees)
DRAFT - DO NOT CITE OR QUOTE
XV
-------
c|)si = secant angle of internal or interface friction for the critical potential
slip surface (degrees), as defined in Figure 6-8
c[)ti = tangent angle of internal or interface friction for the critical potential
slip surface (degrees), as defined in Figure 6-8
DRAFT - DO NOT CITE OR QUOTE
-------
Chapter 1
Introduction
1.1 Overview
1.1.1 Purpose
The guidance provided in this document is designed to be used as a tool to provide design
information to facility owners/operators, engineers, and regulators regarding cover systems for
municipal solid waste (MSW) and hazardous waste (HW) landfills being remediated under the
Comprehensive Environmental Response, Compensation and Liability Act (CERCLA),
Resource Conservation and Recovery Act (RCRA) Corrective Action, and sites regulated under
the RCRA. For sites at MSW and industrial waste (IW) facilities that are subject to State
permits, the technical information contained in this document may be used to supplement
existing guidance in order to achieve compliance with those permits (EPA, 1993; and EPA,
2003). This guidance document provides an update to the previous U.S. Environmental
Protection Agency (EPA) guidance on this subject "Design and Construction of RCRA/CERCLA
Final Covers" (EPA, 199la).
In comparison to the scope of the 1991 EPA cover system guidance document, the scope of this
document has been expanded to address a number of new topics including design criteria
development, new types of geosynthetics (such as geosynthetic clay liners (GCLs)), alternative
materials and designs (including evapotranspiration (ET) barriers and capillary barriers), special
design issues, lessons learned from the closure of existing landfills, performance monitoring of
cover systems, maintenance of cover systems to achieve the required design life, and site end
use. Significant advances in the technology for cover system design and construction have
occurred since 1991. These advances are reflected in this document.
Final cover systems (hereafter referred to as "cover systems") are used at landfills and other
types of waste management units (e.g., waste piles and surface impoundments) to contain waste
and any waste by-products (e.g., leachate or landfill gas), control moisture and air infiltration
into the waste, and prevent the occurrence of odors, disease vectors, and other nuisances. Cover
systems are also used to meet erosion, aesthetic, and other post-closure site end use criteria for
waste management sites. These systems are intended to achieve their functional requirements
for time periods of many decades to hundreds of years.
As illustrated by Figure 1-1, cover systems form one component of the integrated group of
engineered systems used at landfills to protect human health and the environment. Other
components include liners, daily and intermediate covers, leachate collection and removal
systems, gas collection and removal systems, and surface-water management systems.
DRAFT - DO NOT CITE OR QUOTE 1 -1
-------
Gas
Collection and Removal
System
Surface-Water
Management System
Leachate Collection
and
Removal System
Leakage T
Detection System > Composite
Double Liner Bottom Liner1
System
Figure 1-1. Example of Engineered Systems Used at Landfills.
Cover systems are also placed over old dumps as part of the remediation and final closure of
these facilities and over contamination source areas that can be at the ground surface or at
shallow depth. When used for these applications, the cover system may again be one component
of an integrated group of engineered systems used for facility closure or source containment
(Figure 1-2). The cover system components for these facilities are often similar to the
components used to close new landfills. However, as discussed subsequently in this document,
some of the design issues faced in closing dumps or in implementing source containment
remedies at contaminated sites are different from the design issues faced in closing new landfills.
The cover system itself can consist of multiple layers of different types of soils and/or
geosynthetics, each with one or more specific functions. The cover system components are
briefly introduced in Section 1.5 and discussed in more detail in Chapter 2. Although gas
management issues are discussed in Section 1.4 and Chapter 5 of this document, the information
provided on regulatory requirements for MSW landfills under the Clean Air Act (CAA) is
cursory and is not the intended use of this guidance document.
The waste to be contained can be municipal solid waste (MSW), hazardous waste (HW), low-
level radioactive waste, industrial waste (IW), remediation waste, incinerator or coal-combustion
ash, construction and demolition waste (C&DW), sewage treatment or industrial process sludge,
or some other material. The cover system is installed on top of the waste shortly after a specific
landfill cell or unit has been filled to capacity in the case of a new landfill, at the time of site
remediation and closure in the case of an old dump, or at the time of site remediation in the case
of a contaminated site.
DRAFT - DO NOT CITE OR QUOTE
1-2
-------
Leachate and Groundwater
Recovery Well
Cover System
Figure 1-2. Example of Engineered Systems Used at Old Dumps or Contamination
Source Areas.
1.1.2 Classification of Cover Systems
At present, cover system designs are based on one or more of three different principles for
preventing or minimizing water percolation into waste. Each of these is briefly discussed below.
Hydraulic Barrier: This type of cover system uses a low-permeability physical barrier to
impede the downward migration of water into the waste (Figure 1-3). Hydraulic barrier
materials most commonly include compacted clay liners (CCLs), GCLs, geomembranes (GMs),
and combinations of these materials. Other barrier materials (e.g. asphaltic concrete) have also
been used. A hydraulic barrier is generally used with additional cover system components.
However, recently, at a few MSW landfill sites, a GM barrier was used alone as a cover system
(Gleason et al., 1998, 1999, 2001). In many cases and especially on sideslopes, an internal
drainage layer is included above the hydraulic barrier to drain the overlying layers, promote
lateral drainage, and prevent the buildup of hydraulic head in the cover system. A
surface/protection layer is often installed as the topmost layer to protect the hydraulic barrier
from erosion, exposure to wet-dry cycles, exposure to freeze-thaw cycles, biointrusion (intrusion
by plant roots, burrowing animals, and humans), and ultraviolet degradation and for temporary
storage of infiltrating water for subsequent uptake by vegetation, if present. Water movement
through cover systems with hydraulic barriers can occur as either saturated or unsaturated flow,
depending on site-specific conditions (particularly climate). Current EPA regulations and
existing requirements for cover systems at landfills are developed around the use of hydraulic
barriers.
DRAFT - DO NOT CITE OR QUOTE
1-3
-------
ET
Surface/Protection
Layer
Internal Drainage
Layer
Hydraulic
Barrier
-~— Waste -=_
P = Precipitation
ET = Evapotranspiration
R = Runoff
L = Lateral Drainage
Figure 1-3. Hydraulic Barrier Type of Cover System.
ET Barrier: This type of cover system has been developed for use at arid and semi-arid sites.
ET barriers consist of a thick layer of relatively fine-grained soil capable of supporting
vegetation (Figure 1-4). Soil types used for construction of ET barriers include silty sands, silts,
and clayey silts. ET barriers exploit two characteristics of fine-grained soils: (i) high water
storage capacity (i.e., they can store a significant amount of water before gravity drainage: they
have a high field capacity); and (ii) low hydraulic conductivity, even at high degrees of
saturation. High soil water storage capacity allows storage of water that infiltrates the barrier
until it can later be removed by ET. Low hydraulic conductivity limits advancement of the
wetting front into the barrier during seasonal wet periods (rainfall or snow melt). An ET barrier
should be sufficiently thick such that the soil water content does not increase near the base of the
barrier; all changes in soil water storage should occur in the upper portion of the barrier (Figure
1-4). Otherwise, percolation through the cover system can occur. The required barrier thickness
is a function of the frequency and intensity of precipitation, the unsaturated hydraulic properties
of the soil, the type and vigor of vegetative cover, the rate at which water can be removed by ET,
and other factors. Barrier thickness typically ranges from about 0.9 m to more than 2 m. ET
barriers often have a surface layer to support vegetation and provide erosion protection.
DRAFT - DO NOT CITE OR QUOTE
1-4
-------
Fine-Grained '•'
soil •i-.y/.v /::;:•.
Shallow Moisture Content
••' (seasonal increase/decrease.••.:
'•. . ....... due to .; ..• •....• .'•
precipitation/evapotranspiration).
• . ••-' :.|. Deep Moisture Content-'.•;•;•'•'
'" ' •. .-(no seasonal change) ••:.•.'''••.•":
~-^ ~ --_rj waste -~-=^ -^ --=_
P = Precipitation
ET = Evapotranspiration
R = Runoff
9 = Volumetric Moisture Content
z = Depth
Figure 1-4. ET Barrier Type of Cover System and Representative Soil Moisture Content
vs. Depth Profile.
Capillary Barrier: This type of cover system has also been developed for use at arid and semi-
arid sites. Capillary barriers consist of one or more layers of finer-grained soil overlying one or
more layers of coarser-grained soil. The finer-grained soil layer of the capillary barrier has a
higher water storage capacity than a comparable layer at the same depth without the capillary
break (i.e., a free-draining layer such as an ET barrier). Figure 1-5 illustrates the simplest
configuration for a capillary barrier: a single finer-grained (e.g., clayey silt) soil layer over a
coarser-grained (e.g., sandy) soil layer. At low degrees of soil saturation (i.e., at low matric
potential in Figure 1-5), the hydraulic conductivity of the coarser-grained soil is much less than
that of the finer-grained soil. This is the reverse of the condition that occurs at high degrees of
soil saturation. Capillary barriers store infiltrating water in the finer-grained soil until the water
can be removed by subsequent ET. If they are sloped, capillary barriers can also divert
infiltrating water via unsaturated lateral flow in the finer-grained soil (above the soil interface).
Sometimes a "wicking layer" (with intermediate characteristics to the coarser- and finer-grained
layers) is installed between the coarser- and finer-grained layers to convey lateral flow. At high
degrees of soil saturation (e.g., in a humid climate), the capillary effect breaks down and
percolation through the cover system can occur. Like ET barriers, capillary barriers often have a
surface layer to support vegetation and provide erosion protection.
This guidance document focuses primarily on the hydraulic barrier type of cover system with
limited commentary on the other two types provided mainly in Chapter 3. It is noted, however,
that the use of ET and capillary barrier types of cover systems is becoming more common,
particularly in arid and semi-arid regions of the U.S. While these alternative designs can be
adequate for hydraulic control, they should generally not be used without gas containment
DRAFT - DO NOT CITE OR QUOTE
1-5
-------
components at MSW landfill sites where landfill gas collection control are needed to prevent
offsite gas migration and reduce emissions that are of concern to human health and the
environment.
ET
. . : ::::::::::::: Finer-Grained Soil
Coarse-Grained Soil
P = Precipitation
Waste L = Lateral Drainage
ET =
Evapotranspiration
R = Runoff
Coarse-Grained Soil
•j= Fine-Grained Soil
0)
s
Q.
o
'C
+J
re
Water Content (0)
Figure 1-5. Capillary Barrier Type of Cover System and Representative Unsaturated
Hydraulic Conductivity Functions.
DRAFT - DO NOT CITE OR QUOTE 1 -6
-------
1.1.3 Organization of Document
This document is organized into the following sections:
• List of Acronyms (page viii);
• List of Variables and Units (page xi);
• Introduction (Chapter 1);
• Individual Components of Cover Systems (Chapter 2);
• Alternative Design Concepts and Materials (Chapter 3);
• Hydraulic Analysis and Design (Chapter 4);
• Gas Emission Analysis and Collection System Design (Chapter 5);
• Geotechnical Analysis and Design (Chapter 6);
• Lessons Learned (Chapter 7);
• Performance Monitoring (Chapter 8);
• Post-Closure Maintenance and Site End Use (Chapter 9); and
• References (Appendix A).
1.2 Closure Regulatory Requirements
A starting point in understanding closure requirements for landfills or source area containment
for contaminated sites is to become familiar with the regulations governing the landfill or
environmental remediation project. Although generally well understood, the Federal regulations
applicable to cover systems for RCRA and CERCLA projects are briefly reviewed in this section
of the guidance document.
1.2.1 MSW Landfill Cover Systems
Minimum technical requirements for closure of MSW landfills (MSWLFs) regulated under
RCRA Subtitle D are contained in Title 40 of the Code of Federal Regulations, Section 258.60
(40 CFR §258.60). The regulation allows either a minimum criteria cover system or a
performance-based cover system design. The specific requirements of that regulation are as
follows:
"(a) Owners or operators ofallMSWLF units must install a final cover system that is
designed to minimize infiltration and erosion. The final cover system must be designed and
constructed to:
(1) Have a permeability less than or equal to the permeability of any bottom liner system or
natural subsoils present, or a permeability no greater than 1 x 10~5 cm/sec, whichever is less,
and
(2) Minimize infiltration through the closed MSWLF by the use of an infiltration layer that
contains a minimum 18-inches of earthen material, and
(3) Minimize erosion of the final cover by the use of an erosion layer that contains a
minimum 6-inches of earthen material that is capable of sustaining native plant growth.
DRAFT - DO NOT CITE OR QUOTE 1 -7
-------
(b) The Director of an approved State may approve an alternative final cover design that
includes:
(1) An infiltration layer that achieves an equivalent reduction in infiltration as the
infiltration layer specified in paragraphs (a)(l) and (a) (2) of this section, and
(2) An erosion layer that provides equivalent protection from wind and water erosion as the
erosion layer specified in paragraph (a) (3) of this section."
After the foregoing regulations were issued in October 1991, EPA clarified their intent with
respect to the permeability requirement of the prescriptive minimum criteria cover system in 40
CFR §258.60(a)(l). The Agency's clarification was contained in the Federal Register in June
1992, at 57 FR 28628 (EPA, 1992b). According to this clarification, the cover system is
required to have a hydraulic conductivity less than or equal to that of any underlying liner system
or natural subsoils. The purpose of this requirement is to prevent what the Agency calls the
"bathtub" effect, wherein percolation into the landfill exceeds leakage through the liner system,
causing the accumulation of liquid in the facility. The hydraulic conductivity must also be no
greater than 1 x 10"7m/s.
The EPA (1992b) clarification to the minimum requirements for MSW landfill cover systems is
illustrated in Figure 1-6 for: (i) unlined landfills constructed prior to the effective date of Subtitle
D regulations (Figure l-6(a)); (ii) landfills with a CCL beneath the waste (Figure l-6(b)); and
(iii) landfills underlain by a Subtitle D composite liner consisting of a GM upper component and
a CCL lower component (with the CCL having a maximum hydraulic conductivity of 1 x 10"9
m/s) (Figure l-6(c)). While these minimum requirements seem to indicate that less protective
cover systems are allowed at landfills with less protective liner systems, EPA believes that, all
other factors being equal (e.g., comparable hydrogeologic setting, types of waste, etc.), more
protective cover systems should be used at unlined landfills compared to lined landfills to
minimize the percolation of water though the cover systems and, consequently, the formation of
leachate and migration of such leachate from the units.
It should also be noted that the cover systems required by 40 CFR §258.60 regulations do not
represent "complete" designs in the sense that they are based on a permeability design criterion
only and do not address other design criteria. For example, the cover system shown in Figure
l-6(c) does not include a drainage layer above the GM barrier or an adequate thickness of cover
soil to allow sufficient water storage for healthy surface vegetation. As another example, none
of the designs presented in Figure 1-6 have an adequate thickness of soil protection above the
CCL component of the cover system to protect the CCL from freeze-thaw damage for sites
located in northern climates. As a final example, none of the designs addresses the important
matter of landfill gas transmission beneath the cover system.
DRAFT - DO NOT CITE OR QUOTE 1 -8
-------
0-15mX[^
0.45m
(a)
,^T, Surface/Protection
^J J Layer
CCL Barrier
Surface/Protection
31 Layer
CCL Barrier
Surface/Protection
JLayer
Composite Barrier
(C)
1-^- Waste - ~-^T
Figure 1-6. EPA Prescriptive Minimum Criteria Cover Systems for: (a) Unlined MSW
Landfills; (b) MSW Landfills Underlain by a CCL; and (c) MSW Landfills
Underlain by a GM/CCL Composite Liner.
DRAFT - DO NOT CITE OR QUOTE
1-9
-------
1.2.2 Hazardous Waste Landfill Cover Systems
Minimum technical requirements for closure of permitted HW landfills regulated under Subtitle
C of RCRA are contained in 40 CFR §264.310. Analogous requirements for interim status HW
landfills are contained in 40 CFR §265.310. These regulations allow a performance-based cover
system design; no prescriptive design criteria are provided for HW landfills. The specific
requirements of the regulations for permitted landfills are given below:
"(a) At final closure of the landfill or upon closure of any cell, the owner or operator must
cover the landfill or cell with a final cover designed and constructed to:
(1) Provide long-term minimization of migration of liquids through the closed landfill;
(2) Function with minimum maintenance;
(3) Promote drainage and minimize erosion or abrasion of the cover;
(4) Accommodate settling and subsidence so that the cover's integrity is maintained; and
(5) Have a permeability less than or equal to the permeability of any bottom liner system or
natural subsoils present."
Cover system requirements for interim status HW landfills (40 CFR §265.310) differ in several
details from those given above for permitted facilities. EPA (199la) and Koerner and Daniel
(1997) discussed these differences. EPA generally recommends, however, that cover systems for
interim status HW landfills be designed to the same standards as permitted facilities.
0.6 m
O.SmT •'.'•'.•'-';
0.6 m
Surface/Protection
Layer
~~| Drainage
_| Layer
Composite
Barrier
_J== ZI-=r Waste
Figure 1-7. EPA (1989) Recommended Minimum Cover System for HW Landfills.
DRAFT - DO NOT CITE OR QUOTE
1-10
-------
EPA previously issued minimum technology guidance for cover systems that meet the regulatory
requirements of 40 CFR §264.310 (EPA, 1989). The cover system for HW landfills
recommended in the 1989 EPA guidance consists of (Figure 1-7):
• a top layer containing two components: (i) either a vegetated or armored surface layer,
selected to minimize erosion and, to the extent possible, promote drainage off the cover;
and (ii) a protection layer, comprising topsoil and/or fill soil, as appropriate; the
recommended top layer surface slope is 3 to 5%; the 1989 EPA guidance noted that the
top layer soil component should be at least 0.6-m thick, and that a greater thickness may
be required to assure that the underlying hydraulic barrier is below the frost zone;
• a soil drainage layer with minimum thickness of 0.3 m and a minimum hydraulic
conductivity of 1 x 10"4 m/s that will effectively "minimize water infiltration into the
underlying low-permeability barrier" and have a final slope of at least 3% after
settlement and subsidence or a drainage layer consisting of a geosynthetic material with
performance characteristics equivalent to the soil drainage layer; and
• a composite hydraulic barrier consisting of: (i) a GM with a minimum thickness of 0.5
mm; and (ii) a CCL with a minimum thickness of 0.6 m and a maximum hydraulic
conductivity of 1 x 10"9 m/s; the EPA guidance notes that the entire hydraulic barrier
should lie below the frost zone.
The 1989 EPA guidance indicated that optional layers may be used on a site-specific basis.
According to the 1989 guidance, optional layers may include a gas collection layer placed below
the hydraulic barrier, a biotic barrier component of the protection layer, and geosynthetic or soil
filter layers. All of these types of materials are discussed in more detail in Section 1.5 and
Chapter 2 of this document. The 1989 guidance also discussed the use of alternative designs.
This subject too is discussed in Section 1.3 and Chapter 3 of this document. It is also reiterated
that the 1989 document provides guidance on minimum design criteria. On a case-by-case basis,
it may be necessary to provide additional components or capability to a FIW landfill design. For
example, it may be necessary to specify a drainage layer hydraulic conductivity greater than
1 x 10"4 m/s to assure no unacceptable build-up of hydraulic head in the cover system. As
another example, the thickness of the protection layer may need to be greater than 0.6 m to
adequately protect the hydraulic barrier component from freezing weather impacts in some
northern climates.
1.2.3 Solid Waste Landfill Cover System Performance
Both the MSW and FIW landfill regulations cited above specify as a performance criterion
minimization of water percolation into the waste (or, equivalently, minimization of liquids
migration through the landfill by preventing the bathtub effect). EPA is not yet recommending a
design percolation rate for landfill cover systems.
DRAFT - DO NOT CITE OR QUOTE 1-11
-------
1.2.4 CERCLA Site Cover Systems
The blueprint for remediation of CERCLA sites is contained in the National Contingency Plan
(NCP) of 1990 and the Superfund Amendments and Reauthorization Act (SARA) of 1986.
Remediation of these sites often involves installation of a cover system as part of a source
control remedy for a landfill, waste pile or pit, or heavily contaminated area. EPA (1997a)
reported that containment technologies, which typically include some form of cover system,
have been used for approximately 40% of the source control remedies implemented through
1995 at CERCLA sites.
Design requirements for cover systems at CERCLA sites are generally based on the attainment
of applicable or relevant and appropriate requirements (ARARs). ARARs for cover systems
may include RCRA Subtitle C or Subtitle D regulations. EPA (1991a) provides a detailed
discussion of ARARs in the context of CERCLA cover systems.
CERCLA MSW landfills represent a particular subset of CERCLA sites addressed by EPA's
presumptive remedy guidance (EPA, 1993). CERCLA MSW landfills typically contain a
combination of principally MSW and, to a lesser extent, wastes containing hazardous substances.
CERCLA MSW landfills represent approximately 20% of the total number of CERCLA sites in
the United States (EPA, 1991b). The Agency has developed some presumptive remedies using
preferred technologies for common categories of sites, based on historical patterns of remedy
selection for those categories of sites and EPA's scientific and engineering evaluation of
performance data on technology implementation. For CERCLA MSW landfill sites, EPA
generally considers containment as the presumptive remedy (EPA, 1993). Furthermore, the
Agency has identified cover systems as a component of the source containment presumptive
remedy. EPA (1993) provided the following guidance regarding ARARs for CERCLA MSW
landfill presumptive remedies:
"/« the absence of Federal Subtitle D closure regulations, State Subtitle D closure
requirements generally have governed CERCLA response actions at municipal landfills as
applicable or relevant and appropriate requirements (ARARs). New Federal Subtitle D
closure and post-closure care regulations will be in effect on October 9, 1993 (56 FR 50978
and 40 CFR §258). State closure requirements that are ARARs and that are more stringent
than the Federal requirements must be attained or waived.
The new Federal regulations contain requirements related to construction and maintenance
of the final cover, and leachate collection, ground-water monitoring, and gas monitoring
systems. The final cover regulations will be applicable requirements for landfills that
received household waste after October 9, 1991. EPA expects that the final cover
requirements will be applicable to few, if any, CERCLA municipal landfills, since the receipt
of household wastes ceased at most CERCLA landfills before October 1991. Rather, the
substantive requirements of the new Subtitle D regulations generally will be considered
relevant and appropriate requirements for CERCLA response actions that occur after the
effective date."
DRAFT - DO NOT CITE OR QUOTE 1-12
-------
"RCRA Subtitle C closure requirements may be applicable or relevant and appropriate in
certain circumstances. RCRA Subtitle C is applicable if the landfill received waste that is a
listed or characteristic waste under RCRA, and:
1. The waste was disposed of after November 19, 1980 (effective date of RCRA), or
2. The new response action constitutes disposal under RCRA).
The decision about whether a Subtitle C closure requirement is relevant and appropriate is
based on a variety of factors, including the nature of the waste and its hazardous properties,
the date on which it was disposed, and the nature of the requirement itself. For more
information on RCRA Subtitle C closure requirements, see RCRA ARARs: Focus on Closure
Requirements, Directive No. 9234.2-04FS, October 1989 r
The decision of whether MSW or HW landfill cover system requirements are relevant and
appropriate also depends on the level of cover system hydraulic performance that is necessary to
achieve human or ecological receptor exposure point concentrations that produce acceptable
post-remediation human health and ecological risk estimates.
1.2.5 Liquids Management Strategy
EPA policies and regulations for landfill cover systems have evolved within a framework
originally described by the Agency as a "liquids management strategy." The two main goals of
the strategy are: (i) minimizing leachate generation by keeping liquids out of the landfill (or
source area for a CERCLA remediation); and (ii) detecting, collecting, and removing leachate as
it is generated (EPA, 1991c, 1992a). With this liquids management strategy, keeping water out
of the landfill (or source area) becomes a prime performance objective for the cover system. In
fact, EPA has stated (EPA, 1989):
"Thus, the Agency believes that a properly designed and constructed cover becomes, after
closure, the most important feature of the landfill structure. The Agency requires that the
cover be designed and constructed to provide long-term minimization of the movement of
water from the surface into the closed unit. Where the waste mass lies entirely above the
zone of ground-water saturation, a properly designed and maintained cover can prevent, for
all practical purposes, the entry of water into the closed unit, and thus minimize the
formation and migration of leachate.'"
Figure 1-8 illustrates the benefits of cover system installation in reducing leachate generation
rates. This figure shows leachate generation rates for a GM-lined MSW landfill cell through the
period of active cell operation, the closure period, and the first few years of the post-closure
period. The landfill site is located in Pennsylvania and receives approximately 1,000 mm of
precipitation annually, on average (Bonaparte, 1995). The landfill cover system includes a GM
barrier. Monthly average leachate generation rates during the period of cell filling were up to
3,400 Iphd. Rates for the first three years of the post-closure period were only 70 Iphd. The very
significant effect of cover system installation on the rate of leachate generation is apparent.
Figure 1-9 from Othman et al. (2002) shows similar behavior for a group of MSW and HW
landfill cells that have a cover system that includes a GM. On average, leachate generation rates
typically decreased by a factor of four within one year after closure and by one order of
magnitude within two to four years after closure. Six years after closure, leachate generation
DRAFT - DO NOT CITE OR QUOTE 1-13
-------
rates were between 5 and 1,200 Iphd (mean of 180 Iphd). Nine years after closure, leachate
generation rates were negligible. These data show that well designed and constructed cover
systems can be effective in reducing leachate generation rates to very low or near zero values.
O •"-"-"->
"~" 3500-
ro 3000-
g 2500-
fO 2000-
0
0 150°-
0 1000-
1 50°-
CD n
0 0 J
-1 Ju
-w " —
l-88
Active " Cover
Filling Installed
11
Illllilili .._.
Jul-90
Cell Closed
MSW Landfill
(Pennsylvania)
Jul-91 I Jul-92 I Jul-93
Jan-
89
Jan-90
Jan-91
Jan-92
Jan-93
Jan-94
Date
Figure 1-8. Leachate Generation Rates Over Time at a MSW Landfill in Pennsylvania
(from Bonaparte, 1995). (Flow rates are in liters/hectare/day (Iphd).)
"0
a.
ro
ct
g
IE
0)
c
CD
.2
"to
^
ro
34567
Year Since Final Closure
Figure 1-9. Effect of Cover system Installation on Leachate Generation Rates for 12 MSW
Landfill Cells (shown as circles) and 22 HW Landfill Cells (shown as squares)
(from Othman et al., 2002). (Note: flow rates of 0 Iphd are shown as 0.1 Iphd
on this figure.)
DRAFT - DO NOT CITE OR QUOTE
1-14
-------
1.2.6 Design Life
Consistent with the Agency's liquids management strategy, discussed above, the design life goal
for RCRA and CERCLA cover systems is to minimize infiltration into the waste for as long as
the enclosed waste poses an unacceptable risk to human health and the environment. A
distinction should be made between the minimum post-closure care period of 30 years given in
RCRA regulations and the design life of the cover system. The latter is much longer than 30
years and is defined primarily by the service life characteristics of the material used to construct
the cover system. The service life of CCLs protected from freeze-thaw and other environmental
effects, and not subjected to excessive differential settlements, should be indefinitely long
(Mitchell and Jaber, 1990). The service life of any GM component of the cover system is
dependent on the specific material used and how well the material is protected. The most
extensive service life data currently available are for high density polyethylene (HDPE) GMs.
The data indicate that the service life for commercially-available HDPE GMs will be measured
in terms of at least several hundred years (Hsuan and Koerner, 1998; Hsuan and Koerner, 2002).
Other types of GMs may have different service lives from that for HDPE GMs. Great care
should be used in specifying GM materials to require products that, through polymer type,
additive (e.g., antioxidant) packages, physical robustness, etc., are capable of achieving as long a
service life as possible.
Achieving a design life measured in terms of hundreds of years requires more than just the
selection of durable materials of construction. The design itself should be developed to achieve
the design life criteria. This involves developing a design with adequate slope stability factors of
safety, adequate flow capacity for the internal drainage system, adequate surface-water runoff
controls, adequate freeze-thaw protection, adequate resistance to surface erosion, and appropriate
vegetation or other surface treatment. Many of these design topics are addressed in subsequent
chapters of this document. Recognizing the dynamic nature of the ecosystem in which cover
systems function, post-closure monitoring and maintenance are important elements in achieving
the required design life. Long-term maintenance with respect to surface erosion, biointrusion,
and plant succession (i.e., grasses to shrubs to trees) are particularly important issues in
addressing the design life of a cover system. Monitoring of cover systems after closure is
necessary to both satisfy regulatory requirements and assure the performance of the cover
system. While performance monitoring is important for all closed facilities, it is particularly
such for closed sites, such as old dumps and contamination source areas, and for sites with
alternative cover systems. Monitoring of infiltration, soil moisture, gas emissions, and
settlement is discussed in Chapter 8. The cover system should generally be inspected and
maintained to assure adequate performance of the site in the long term and to comply with
regulatory requirements. Cover system maintenance is discussed in Chapter 9.
1.2.7 Other Regulatory Requirements
In addition to the regulatory requirements cited above, other regulatory requirements may be
ARARs to a landfill closure or CERCLA remediation project. These additional requirements
must be considered on a case-by-case basis. It is essential for proper design and legal
compliance of the project that all potentially applicable regulations be identified during the
design criteria development phase of the project (see Section 1.6 of this document). Other
potential regulatory requirements or ARARs may include:
DRAFT - DO NOT CITE OR QUOTE 1-15
-------
• State-mandated cover system regulations that impose additional requirements beyond the
minimum technical requirements of EPA;
• requirements imposed by other regulations for specific types of wastes, regulated under
the Toxic Substances Control Act (40 CFR §700), such as polychlorinated biphenyls
(PCBs), or Uranium Mill Tailings Remediation Act (40 CFR §192);
• State or Federal (including Federal Emergency Management Agency) requirements for
site surface-water management, landfill gas management, seismic design, or other
requirements that could influence the design of the cover system;
• provisions for management, treatment, and/or discharge of stormwater runoff, leachate,
gas condensate, or other liquids under provisions of the Clean Water Act, including
National Pollution Discharge Elimination System (NPDES) requirements (40 CFR §122)
and proposed landfill point source effluent limitation guidelines (40 CFR §445);
• State requirements for maximum allowable soil erosion rates, erosion control structure
design and performance, and surface-water management structure design and
performance;
• Federal or State requirements for siting, including limitations on construction in
floodplains, disturbance of wetlands, and construction on or near Holocene faults.
1.3 Alternative Design Concepts and Materials
RCRA regulatory requirements provide flexibility for innovation and alternatives by limiting the
use of specific minimum design specifications as much as possible, by providing performance
criteria in lieu of design specifications, and/or by providing administrative procedures for
gaining approval of waivers from RCRA regulatory requirements. Also, under CERCLA
§121(d), ARARs may be waived (refer to guidance).
EPA is open to considering alternative designs on a case-by-case basis. Determinations on the
acceptability of alternative designs for HW landfills are the responsibility of the Regional
Administrator. Statutory requirements must be satisfied by any approved alternative. This
document provides guidance on several of the alternative design approaches and materials that
the Agency may consider on a case-by-case basis. It is anticipated that new design approaches
and materials expect to be considered by EPA in the future as the performance of these
alternatives is demonstrated and proven. As an example of an alternative design, Region 1 of
EPA has issued alternative minimum technology guidance for closure of unlined HW RCRA
landfill sites in that region. The rationale and technical analyses supporting the Region 1
alternative minimum technology guidance is given in EPA (2000a). It is noted that, in Region 1,
this type of landfill often has relatively steep sideslopes (i.e., greater than 6 horizontal:! vertical
(6H: IV)) and soils suitable to construct a hydraulic barrier may not be locally available. A
comparison of the minimum technology guidance from EPA (1989) and EPA (2000a) is
presented in Figure 1-10. Other types of alternative designs may involve ET or capillary barriers
as discussed in Section 1.1 of this document. Alternative design concepts and materials are
discussed in more detail in Chapter 3.
DRAFT - DO NOT CITE OR QUOTE 1-16
-------
The minimum technical requirements for cover systems were developed by EPA to achieve the
liquids management strategy goal previously described. These requirements still represent the
Agency's preferred approach for most types of landfills under most situations. In recent years,
however, the Agency has begun to consider other management strategies for landfill facilities.
Potential strategies include, for example, landfill leachate recirculation and bioreactors (EPA,
1995). EPA believes that new landfill management strategies may lead to new alternative cover
system designs and materials. The Agency is currently considering these types of alternatives on
a case-by-case basis.
0.6m
Soil
•Filter Layer
„., T[- '•'••'.• ':'.'.•'• Granular Soil or Geosynthetic:':'.;.'-i.:-.--'-.'-J
I r-'.v'-; •'.•:'•'.•. ::-•••: 0:'•. •.!;•••'.'•.;• ••v.
0.6m
•Geomembrane(0.5mm)
Low Permeability Soil
(k< 1 x10"9m/s)
Surface/
Protection
Layer
Drainage
Layer
UCIIIICI
1
0.6 m
0.3 m
0.3 m
;
!
Soil
•"• '• •/ . : '.' •' • Granular Soil or Geosynthetic : '.• '.-•'••.'•
'•'.•v;-: •;•::••.-.;' (k>io-3m/s) •.••.:,-.•.• :;.-.;:-:-x'.
' / / / / Low Permeability Soil / / /
////, (k<1_x10-8m/s) ////
Foundation Layer/Intermediate Cover
m
//
Foundation Layer/Intermediate Cover
"Waste
Waste
(a)
(b)
Figure 1-10. Comparison of Cover Systems for HW Landfills: (a) EPA (1989)
Recommended Minimum Technology Cover System; and (b) Region 1
Alternative Minimum Cover System.
The use of monitored natural attenuation is recognized by EPA as a viable technique for
remediation of soil and groundwater at certain sites (EPA, 1999a). The term "monitored natural
attenuation" refers to the reliance on natural attenuation processes to achieve site-specific
remedial objectives within a time frame that is reasonable, compared to that offered by other
more active remediation methods. The "natural attenuation processes" that are at work in such a
remediation approach include a variety of physical, chemical, and/or biological processes that,
under favorable conditions, act without human intervention to reduce the mass, toxicity,
mobility, volume, or concentration of contaminants in soil or groundwater. These in-situ
processes include biodegradation, dispersion, dilution, sorption, volatilization, radioactive decay,
and chemical or biological stabilization, transformation, or destruction. EPA is aware of
situations where monitored natural attenuation has been proposed along with a permeable (e.g.,
granular) cover system as a source control remedy for a CERCLA landfill. Since this approach
is not consistent with the Agency's liquids management strategy, EPA will evaluate these cover
DRAFT - DO NOT CITE OR QUOTE
1-17
-------
systems very carefully on a case-by-case basis and, in some cases, will require that an in-situ
treatment technology be used with this approach to complement natural attenuation and a
demonstration of the technical practicability of the technology. As an example, this remediation
approach is being used by EPA Region 1 for the Somersworth Sanitary Landfill Superfund Site.
As outlined in the 1995 Consent Decree for the site, the Preferred Source Control Remedy
includes:
• "placement of a permeable cover over the landfill allowing precipitation to flush
contamination from the waste management area. This cover will remain as long as
contaminants continue to leach from the waste within the waste management area and
the chemical treatment "wall" is functioning. After the Final Cleanup Levels have been
achieved and can be maintained with use of the treatment "wall, " an appropriate landfill
cover to close the landfill that is consistent with the ROD (Record of Decision) shall be
installed and maintained.";
• "installation of a treatment wall composed of impermeable barrier sections and
permeable, chemical treatment sections to provide in-situ (in-place), flow-through
treatment of contaminated ground water at the down-gradient edge of the waste
management area." (the site and the pilot-scale treatment wall is described in EPA
(1999b)); and
• "enhancements ... and additional source control measures, if 'necessary".
The Consent Decree for this site places the burden of using the alternative source control remedy
on the party implementing the remedy. If the Preferred Source Control Remedy does not meet
the specified performance standards, a Contingent Source Control Remedy, including installation
of a cover system that meets RCRA Subtitle C requirements and other ARARs, may need to be
implemented.
1.4 Gas Management Requirements
Landfill gas collection and control is necessary at some MSW landfills, a limited number of HW
landfills, and some CERCLA remediation sites. Most modern MSW landfills built to current
regulatory standards have landfill gas collection and control systems. Some sites recover the gas
for its energy potential, which may help to offset regulatory compliance costs. As of January
1999, there were about 300 MSW landfill gas-to-energy projects active in the U.S and several
hundred more planned or in construction (Thorneloe, 2000).
Anaerobic decomposition of organic material in waste is the principal source of landfill gas and
a significant cause of settlement of the waste mass. Some industrial wastes, however, can
generate gas by inorganic chemical reactions. Gas production rates vary with the composition
and age of waste, waste volume, waste moisture content, and other factors. MSW landfill gas
consists mainly of methane and carbon dioxide, with lesser concentrations of nitrogen, oxygen,
sulfides, ammonia, and other constituents, and trace concentrations of a variety of volatile
organic compounds, including vinyl chloride, ethylbenzene, toluene, and benzene
(Tchobanoglous, 1993). Landfill gas can be a significant threat to human health and the
environment (EPA, 2000). Because of this, CAA regulations establish requirements for MSW
DRAFT - DO NOT CITE OR QUOTE 1-18
-------
landfill gas collection and control at certain facilities. Gas generation in a MSW landfill cell can
extend over a period of 25 years or more, or gas generation can be accelerated through the use of
leachate recirculation. An idealization of gas generation rates in MSW landfills is presented in
Figure 1-11.
620
TO
(D
I
*^
CD
'31C
g
'•6
2
CL
w
TO
O
-Gas production from a landfill
with adequate moisture to support
complete anaerobic digestion
of the organic fraction of MSW
Gas production from the same landfill
with inadequate moisture to support
complete anaerobic digestion
10 15
Time (years)
20
25
Figure 1-11. Idealization of Gas Generation Rates in MSW Landfills (from
Tchobanoglous, 1993).
Gas emissions from MSW landfills are presently governed by two sets of regulations that may
influence the design of landfill gas collection and control systems associated with the cover
systems. A third regulation was proposed in November 2000 (EPA, 2000). RCRA Subtitle D
regulations address the personal and fire/explosion safety aspects of landfill gas under 40 CFR
§258.23, which requires:
"(a) Owners or operators ofallMSWLF units must ensure that:
(1) The concentration of methane gas generated by the facility does not exceed 25 percent of
the lower explosive limit for methane in facility structures (excluding gas control or recovery
system components); and
(2) The concentration of methane gas does not exceed the lower explosive limit for methane
at the facility property boundary."
The second set of regulations governing MSW landfill gas is the New Source Performance
Standards (NSPS) and Emissions Guidelines (EG) promulgated under the Clean Air Act (CAA).
The NSPS and EG regulate emissions of non-methane organic compounds (NMOCs) as a
surrogate to total landfill gas emissions (40 CFR §60.755). MSW landfills with design
capacities equal to or greater than 2.5 million megagrams and 2.5 million cubic meters with
NMOC emission estimates of 50 megagrams or more per year must have: (i) a well designed and
DRAFT - DO NOT CITE OR QUOTE
1-19
-------
operated gas collection system; and (ii) a control system device capable of reducing NMOC
mass in the collected gas by 98%.
The third regulation proposes national emission standards for hazardous air pollutants
(NESHAP) for MSW landfills identified as major sources of hazardous air pollutants (HAP)
listed in Section 112(b) of the CAA and some MSW landfills identified as area sources (EPA,
2000). The proposed NESHAP contains the same requirements as the EG and NSPS as well as
some additional requirements to further reduce HAP emissions to the level reflecting the
maximum achievable control technology (MACT). The total impact on MSW landfills is
expected to be limited.
Additional information on gas management regulations for MSW landfills can be found at
http ://www. epa. gov/ttn/atw/landfill/landflpg.html.
Waste-generated gas affects cover systems in several ways. The presence or absence of gas
influences the selection of the type of hydraulic barrier material. GMs are generally better
barriers to gas migration than soils, with the possible exception of intact CCLs at or near
saturation (although low-permeability soils at saturation can have low shear strength and drying
of the soil is a concern). Also, it may be necessary to install a gas collection layer beneath the
barrier to convey gas to outlets through the cover system, or alternatively to install gas extraction
wells or trenches at a sufficiently close spacing to prevent gas build-up beneath the barrier.
A factor sometimes overlooked in the closure of old landfills and in remediation of
contamination source areas is that placement of a cover system will trap any gas being generated
by the waste. Gas generation rates at these facilities may be slow enough that gas generation is
not even recognized as a design issue. Yet after cover system installation, gas pressure can
slowly build up. This process may eventually lead to one or more of the following: (i) problems
with cover system performance, including a reduction in the factor of safety along interfaces in
the cover system below the hydraulic barrier; and (ii) for unlined or inadequately lined landfills
and contamination source areas, subsurface gas migration. Subsurface gas migration has caused
adverse groundwater quality impacts at many older, unlined landfills and may also cause
increases in atmospheric emissions of gas and safety or health impacts to nearby residences both
from gas migrating through the soil and being released from groundwater passing beneath a
residence. EPA recommends that the potential for landfill gas generation and impacts to nearby
residences or businesses always be carefully evaluated as part of any landfill closure or
remediation project.
Another factor to be noted is the trend towards recirculation of leachate and addition of other
liquids to promote decomposition of landfilled waste. This results in faster and greater gas
production than conventional landfills. These factors need to be considered in the final cover
design to ensure adequate protection to near-by residents and the environment.
1.5 Typical Cover System Components
Components of a typical hydraulic barrier cover system are briefly introduced here and discussed
in more detail in Chapter 2. The usual sequencing of these components is illustrated in Figure
DRAFT - DO NOT CITE OR QUOTE 1 -20
-------
1-12. Depending on a site's regulatory status, not all components listed below may be required
as part of the final cover system.
1.5.1 Surface Layer
The topmost component of a cover system is the surface layer. The primary functions of this
layer are to resist erosion by water and wind, be maintainable, and provide a growing medium
for vegetation, if present. The surface layer may also serve other purposes, such as promoting
ET or satisfying project aesthetic, ecological, or end use criteria.
Materials that can be used for cover system surface layers include: (i) topsoil; (ii) amended
topsoil; (iii) gravel-soil mixtures; (iv) gravel; (v) riprap; (vi) articulated block systems; (vii)
asphaltic concrete; and (viii) other materials. Of these materials, topsoil is, by far, the one most
commonly used. Suitable topsoil promotes growth of vegetation, thereby maximizing the ET
component of the cover system water balance. Vegetation also decreases the quantity and
velocity of stormwater runoff on the cover system slopes and reinforces the topsoil; both of these
effects reduce the rate of topsoil erosion in comparison to a topsoil layer without vegetation. At
sites with conditions unfavorable to maintaining an adequate growth of vegetation (e.g., sites
with steep slopes or in semi-arid or arid climates), gravel-soil mixtures, gravel, riprap, articulated
block systems, or other materials may be used for the surface layer.
1.5.2 Protection Layer
A protection layer may serve several functions:
• protect underlying layers from erosion;
• protect underlying layers from exposure to wet-dry cycles, which may cause degradation
of these layers;
• protect underlying layers from exposure to freeze-thaw cycles, which may cause
degradation of these layers;
• serve as a barrier to human, burrowing animal, or plant root intrusion (i.e. a biobarrier);
DRAFT - DO NOT CITE OR QUOTE 1 -21
-------
Surface Layer'
v- ';\-{:^f-:y;:'^^:-:: Protection'
'.'./•.:'.• :'• • Drainage Layer '•_ •
Gas Collection Layer
Foundation Layer
Figure 1-12. Typical Hydraulic Barrier Cover System Components.
• temporarily store water that has infiltrated through the surface layer until the water
returns to the atmosphere through ET; this action provides a water reservoir to support
plant growth and reduces infiltration into underlying cover system layers; and
• restrict emissions of radon gas for those wastes, such as uranium mill tailings, that emit
radon.
On-site or locally available soil is usually suitable for protection layer construction if the primary
functions of the layer are to support vegetation and protect underlying layers from cracking due
to wet-dry and freeze-thaw effects. However, if the primary role of the protection layer is to
prevent biointrusion, cobbles, asphaltic concrete, or similar materials are typically required.
1.5.3 Drainage Layer
In a hydraulic-barrier type cover system, a drainage layer may be required beneath the protection
layer and above the hydraulic barrier, particularly on sideslopes. A drainage layer may serve
several functions:
• limit the buildup of hydraulic head on the underlying hydraulic barrier, which minimizes
percolation of water through the barrier;
DRAFT - DO NOT CITE OR QUOTE
1-22
-------
• drain the overlying surface and protection layers, which increases the available water-
storage capacity and helps to minimize erosion by reducing the time during which the
layers remain saturated with water; and
• reduce the seepage forces in the protection, surface, and drainage layers, which improves
cover system slope stability.
Materials used for drainage layers include sand, gravel, geotextile (GT), geonet (GN), and
geocomposite (GC) drainage materials. The material used should have adequate hydraulic
conductivity to minimize the buildup of hydraulic head above the hydraulic barrier and adequate
hydraulic transmissivity to convey the design flow rate. If gravel or a GN is used for the
drainage layer, a filter layer will usually be needed between the drainage layer and the overlying
protection layer to prevent fines from clogging the drainage layer. GT filter layers are typically
used to achieve this function, although soil filter layers can also be used. If the drainage layer
consists of gravel, and the underlying barrier is a GM, a GT cushion layer is typically needed
between the GM and gravel. One of the most important aspects of designing a drainage layer is
providing for free drainage at the drainage layer outlet.
1.5.4 Hydraulic Barrier
The function of the hydraulic barrier is to minimize percolation of water through the cover
system by impeding infiltration into the barrier and by promoting storage or lateral drainage of
water in the overlying layers. Properly designed, constructed, and maintained hydraulic barriers
can virtually eliminate infiltration into the waste. Hydraulic barriers also restrict migration of
gas or volatile constituents from the waste mass to the atmosphere.
Materials used for hydraulic barrier construction include GMs, GCLs, and CCLs. Each of these
barrier materials may be used alone or in combination. It has been shown, however, that, all else
being equal, a cover system with a composite barrier consisting of GM/CCL, GM/GCL, or
GM/GCL/CCL allows less percolation than a cover system with a GM, GCL, or CCL barrier.
1.5.5 Gas Collection Layer
Gas collection layers may be necessary beneath cover system barriers for wastes that generate
gas or emit volatile constituents. These layers are designed to have adequate in-plane gas
transmissivity to convey gas to passive gas vents, active gas wells, or trenches. Gas collection
layers are typically a necessary complement to systems that utilize passive gas vents. Gas
collection layers may not always be needed for landfills with active gas extraction systems,
depending on gas generation rates in the landfill, extraction well spacing, presence or absence of
horizontal gas trenches, and other factors.
Gas collection layers may be constructed of granular materials (e.g., sand or gravel) or
geosynthetics (e.g., GT, GN, GC). The selected material must have adequate transmissivity to
minimize the build up of gas pressures beneath the barrier and convey the design gas flow rate.
When a granular material is used, a separation layer (typically a GT) may be needed to separate
the granular material from the overlying barrier.
DRAFT - DO NOT CITE OR QUOTE 1 -23
-------
1.5.6 Foundation Layer
The foundation layer is the bottom-most component of the cover system. The functions of the
foundation layer are to provide grade control for cover system construction, adequate bearing
capacity for overlying layers, a firm subgrade for compaction of overlying layers, a smooth
surface for installation of overlying geosynthetics, and, in some applications, a buffer zone to
reduce the potential effects of waste differential settlements on the cover system components.
Materials most often used for the foundation layer include on-site or locally available soils.
Sometimes, intermediate cover soil already in place is used for all or a portion of the foundation
layer. In a few situations, waste material can be used to construct the foundation layer. If
constructed of granular material, the foundation layer may also serve as a gas collection layer.
1.6 Design Criteria Development and Conceptual Design
1.6.1 Overview
Gross et al. (2002) present the results of a survey conducted for EPA on problems and lessons
learned at representative landfill facilities located throughout the U.S. The survey identified 69
modern landfill facilities that had experienced 80 liner system or cover system problems. For the
study, a "modern facility" was considered one with components substantially meeting current
EPA regulations and constructed and operated to the U.S. state-of-practice from the mid-1980's
forward. Almost 30% of the problems identified in the study involved landfill cover systems.
The percentage of cover system problems for the 69 facilities will likely be higher in the future
since a number of these facilities were active and did not yet have a cover system. The primary
factor contributing to the cover system problem in most cases was inadequate design.
The number of facilities in the EPA study is small compared to the total number of modern
landfills nationwide. However, the search for problem facilities was not exhaustive. The
Agency believes many more facilities than identified in the study have experienced the types of
problems identified in the study. As pointed out in the EPA study, the single factor that can most
improve the performance record for waste containment systems is improved design practice by
the engineering community. It is imperative and consistent with the standard of professional
care that engineers prepare complete, detailed, and proper designs of cover systems. Simple and
incomplete design approaches intended to simply satisfy regulatory requirements and "get the
grass growing" are not acceptable. This guidance document is intended to contribute to
improved practices with respect to the design of cover systems.
The critical first steps in designing a landfill cover system involve: (i) developing the criteria that
will be used to guide the design; (ii) preparing a conceptual design using these criteria; (iii)
identifying data gaps based on the conceptual design; and (iv) performing predesign studies to
generate the data needed to prepare the detailed design and construction plans and specifications.
Design criteria development is addressed in more detail below.
DRAFT - DO NOT CITE OR QUOTE 1 -24
-------
1.6.2 Regulatory Requirements
The first step in establishing design criteria is to identify all applicable regulatory requirements
(for a RCRA Subtitle D or Subtitle C facility) or ARARs (for a CERCLA site remediation).
General guidance on applicable regulatory requirements was given in Section 1.2 of this
document. Federal regulations are found in the Code of Federal Regulations and are available on
the U.S. Government Printing Office website at http://www.access.gpo.gov/nara/cfr/cfr-table-
search.html. State and local regulations may also be available on-line.
1.6.3 Climatic Criteria
Climate significantly affects cover system design and performance. For example, the typical
approach to preventing water percolation through a cover system for a facility in the eastern U.S.
is to use a low-permeability hydraulic barrier. In arid regions of the western U.S., however, the
same design objective can be achieved using an ET barrier. As another example, climatic factors
influence the thickness of the cover soil required to protect an underlying hydraulic barrier from
the effects of freeze-thaw. Further, climatic factors greatly affect the types of vegetation that can
be grown on a cover system.
Climatic criteria to consider in the design of a cover system include the amount and seasonal
distribution of precipitation, duration of specific storm events (e.g., 1-hour storm event, 24-hour
storm event, etc.), intensity of specific storm events (e.g., 25-year recurrence interval storm
event, 100-year recurrence interval storm event, probable maximum precipitation (PMP), etc.),
seasonal temperature variations and extremes, depth of frost penetration, quantity of snow melt,
wind speed and direction, solar radiation and humidity. In some areas (e.g., cold, arid), the
controlling climatic criterion for percolation may be snowmelt.
1.6.4 Physical and Engineering Criteria
Physical criteria that should be considered in designing a cover system include the lateral limits
of waste, property setback requirements, if any, height of facility above surrounding ground,
sideslope length and inclination, top deck length and inclination, depth of waste within the
facility, type and thickness of interim cover, and potential for the waste to generate gas. A
distinction should be made at this stage between proposed landfills where the design engineer
has control over essentially every physical parameter for the facility versus an existing landfill or
CERCLA remediation site where the design engineer starts the design process by considering
the pre-existing site conditions. The consequences of a certain design action can be quite
different for these two situations. For example, it is a relatively straightforward matter to design
and construct a stormwater management or slope stability terrace or bench for a new landfill.
Conversely, design and installation of a terrace or bench for a cover system on a steep pre-
existing landfill slope can be difficult or infeasible. The latter type of design requires either that
a cut be made into the existing waste slope or, alternatively, that the terrace be built up above the
waste slope using soil fill and foundation/slope reinforcement techniques. These kinds of
differences should be considered by the design engineer during design criteria development.
Design criteria development for a cover system should also consider a number of engineering
criteria. The design engineer should carefully consider which engineering criteria are relevant
for a particular facility, and then apply them appropriately. For each engineering criterion that
DRAFT - DO NOT CITE OR QUOTE 1 -25
-------
must be satisfied, the design engineer should define: (i) performance requirements for that
criterion; (ii) method of analysis or evaluation; and (iii) required input parameters with
numerical values for each parameter and at least qualitative, if not quantitative, consideration of
the uncertainity (i.e., standard deviation, standard error, etc.) associated with the selected
numerical values. As an example, a common engineering criterion for landfill cover systems is
long-term static stability of the waste mass beneath the cover.
Table 1-1. Common engineering criteria for RCRA and CERCLA cover systems.
Slope Stability
• Foundation stability
• Waste mass stability
• Cover system veneer stability
• Pseudo-static stability analysis
• Other stability conditions
Settlement (Total and Differential)
• Foundation total settlement
• Waste mass total settlement
• Foundation differential settlement
• Waste mass differential settlement
Seismic Deformation Analysis
• Foundation liquefaction
• Waste mass deformation
• Cover system deformation
Surface-Water Runoff Control
• Estimated peak flow rate
• Surface-water control structure design
(benches, channels, and retention
ponds)
Geosynthetic Component Performance
• GT filter layer requirements
• GT clogging potential
• GN/GC flow rate
• GN/GC clogging potential
• GN/GC compression resistance
• GN/GC outlet design
• GT cushion layer requirements
• GM design (type, thickness, elongation
and strength requirements)
• GCL design (type, internal
reinforcement, overlap)
Erosion Control and Vegetation
• Rill and interrill erosion
• Gully formation (tractive force analysis,
critical distance for gully formation,
permissible velocity analysis)
• Wind erosion
• Vegetation requirements (type,
planting, fertilizer, amendments)
• Temporary and permanent erosion
control material requirements
Soil Component Performance
• Erosion resistance of surface layer
• Biointrusion resistance
• Water storage capacity
• Frost penetration depth
• Drainage layer flow rate
• Drainage layer clogging potential
• Drainage layer outlet design
• Granular filter layer requirements
• Soil barrier hydraulic design (suitable
soil availably, thickness, hydraulic
conductivity)
Hydraulic Performance
• Cover system water balance
• Percolation through cover system
• Water flow in drainage layer
• Maximum head in drainage layer
Gas Emission Control
• Gas emission rate analysis
• Gas flow and pressure in collection
layer
• Gas collection system (active or
passive)
• Gas treatment requirements
DRAFT - DO NOT CITE OR QUOTE
1-26
-------
The performance requirement for this criterion is usually expressed in terms of a factor of safety
against slope instability. The minimum acceptable factor of safety might be 1.5, for example.
(A discussion of recommended slope stability factors of safety is given in Chapter 6 of this
document.) A method of analysis that could be used to evaluate this criterion is a two- or three-
dimensional limit equilibrium method of slices. Input parameters for the evaluation include the
geometry of the waste, unit weight and shear strength of the waste, existence of any perched or
continuous zones of leachate in the waste, existence of landfill gas pressures beneath the cover
system, and the thicknesses, unit weights, internal shear strengths, and interface shear strengths
of the cover system installed over the waste.
A partial list of engineering criteria that are frequently considered in the design of RCRA or
CERCLA cover systems with the components shown in Figure 1-12 are listed in Table 1-1. Not
all criteria apply to all cover systems. The foregoing list of criteria, while extensive, is by no
means exhaustive. Additional criteria will need to be considered on a case-by-case basis. Also,
particular attention should be given to applicable engineering criteria any time an alternative or
innovative cover system is proposed, as past precedent for such systems will, by definition, be
limited or nonexistent.
1.6.5 Aesthetic and Land Use Criteria
Aesthetic and land use criteria are becoming more important in the design of cover systems.
More and more, facility owners, regulators, and the local community are sensitive to the
aesthetics of closed waste management sites. Today, it is not uncommon to design aesthetic
enhancements into site closure projects. When such enhancements are to be used, they should be
adequately designed in their own right, and any impact they may have on any other engineering
criterion identified previously in this section should be addressed. Examples of aesthetic
enhancements that have been incorporated into cover systems include:
• construction of an undulating sideslope to provide a more natural looking landform
(compared to long, planar sideslope);
• planting of trees and shrubbery on terraces; and
• construction of decorative block retaining walls.
Increasingly, beneficial post-closure land uses are being considered in the design of cover
systems for landfill closures and CERCLA remediations. The most common types of end uses
are parks, hiking trails, sports fields, and golf courses. The selected end use can have a
significant impact on cover system design. For example, if a site is to be used for a golf course
or other facility with a vegetated surface layer that requires irrigation, the cover system may
require an internal drainage layer and a barrier that includes a GM to control percolation through
the cover system (Hauser, 2000). Figure 1-13 shows a completed CERCLA remediation in
southern California where the site was closed with a multi-component soil and geosynthetic
cover system, and a golf course was developed on the cover system. Further discussion of
aesthetic and post-closure land uses for cover systems is given in Chapter 9 of this document.
DRAFT - DO NOT CITE OR QUOTE 1 -27
-------
1.6.6 Ecological Criteria
Conventional engineering approaches for designing cover systems often fail to fully consider
ecological processes at work in the local environment. Natural ecosystems effective at capturing
and/or redistributing materials in the environment have evolved over millions of years.
Consequently, when contaminants are introduced into the environment, ecosystem processes
begin to influence the distribution and transport of these materials, just as they influence the
distribution and transport of nutrients that occur naturally in ecosystems (Hakonson et al, 1992).
As the ecological status of a cover system changes, so will performance factors such as water
infiltration, water retention, ET, soil erosion, and biointrusion. An objective often overlooked in
designing cover systems is to cause subsequent ecological change to enhance and preserve the
encapsulating system. Only through a holistic ecological approach can long-term maintenance
requirements for cover systems be truly minimized (Caldwell and Reith, 1993). Consideration of
natural analogs can enhance the design of a cover system by disclosing those processes that are
active in a given environment or the mechanisms that could lead to failure. These mechanisms
can then be avoided through appropriate design and construction. Natural analog studies provide
clues from past environments that can be applied to the long-term behavior and performance of a
cover system. Analog studies involve the use of logical analogy to investigate natural and
archaeological occurrences of materials, conditions, or processes that are similar to those known
or predicted to occur in some part of the cover system (Waugh, 1994). Perhaps the simplest
examples of a natural analog for a cover system are the stable soil geomorphology in the locality
of a project. Local soil geomorphology may be an indicator of the erosional stability of local
soils used for the surface/protection layer in a cover system. For example, if a local glacial till is
to be used for the surface/protection layer of a landfill, and all the local landforms containing
that fill have evolved with slopes no steeper than a certain value, then use of that till on steeper
cover system slopes contravenes the local geomorphological evidence, suggesting a greater
likelihood of long-term maintenance requirements than might otherwise be the case.
A primary goal of design is to achieve a cover system that is as maintenance-free as possible.
While it is debatable as to whether the need for all long-term maintenance can be eliminated,
significant progress is possible with respect to current engineering practice. Moreover, in
virtually all cases, some degree of maintenance or post-construction refinement may be
necessary until the cover system reaches a state of equilibrium with its inherent environment.
An important point often not recognized is that a cover system should be stabilized with
vegetation comprising plant communities that closely emulate a selected local "climax"
community (Caldwell and Reith, 1993). A climax community, in ecological terms, is defined by
the environmental parameters of the community (e.g., climate, soil, and landscape properties,
fauna, and other flora). Central to the concept of "climax" is the community's relative stability
in the existing environment (Whittaker, 1975). A diverse mixture of native plants on a cover
system maximizes water removal through ET (Link et al., 1994). The cover system is then more
resilient to natural and man-induced catastrophes and fluctuations in environments. Similarly,
biological diversity in cover system vegetation is important to community stability and resilience
given variable and unpredictable changes in the environment resulting from pest outbreaks,
disturbances (overgrazing, fires, etc.), and climatic fluctuations.
DRAFT - DO NOT CITE OR QUOTE 1 -28
-------
Figure 1-13. Example of Post-Closure Land Use: Closed California CERCLA Site Used as
a Golf Course.
fig 1-13
Local native species that have been selected over thousands of years are best adapted to
disturbances and climatic changes (Waugh, 1994). In contrast, planting of non-native species, as
is common in the current standard-of-practice for landfill and containment system engineering, is
genetically and structurally monotonous (Harper, 1987) and therefore more vulnerable to
disturbances. Pedogenic processes gradually change the physical and hydraulic properties of
earthen material cover systems (Hillel, 1998). Plant communities inhabiting the cover system
will also change in response to these changes in soil properties.
A cover system that is to last for hundreds of years, or longer, should be designed as an integral
component of a larger dynamic ecosystem. Cover system components initially designed for a
specific purpose such as a barrier or drainage layer will not function independent of one another.
Therefore, these systems should be considered not only individually, but also as a system (linked
assemblage of components). Inevitable changes in physical and biological conditions should be
taken into account to help ensure the long-term effectiveness of the cover system.
DRAFT - DO NOT CITE OR QUOTE
1-29
-------
Chapter 2
Individual Components of Cover Systems
2.1 Introduction
As described in Section 1.5, a typical hydraulic barrier cover system will have the following
components: surface layer, protection layer, drainage layer, hydraulic barrier, gas collection
layer, and foundation layer (Figure 1-12). Not all components are necessary for all cover
systems. For example, a gas collection layer is unnecessary if the underlying waste does not
generate gases that require collection or control. Each component in a cover system serves a
specific purpose and must function for its intended design life. For instance, the gas collection
layer facilitates collection and control of decomposition gases or vapors generated by the waste
or remediation source area material and must function as long as the gases or vapors are
produced. The components of a cover system should interact as a system. The gas collection
layer, for example, works properly only if one of the overlying layers (typically the hydraulic
barrier) serves as a barrier to gas migration, allowing the gases to accumulate in the gas
collection layer, where they can be removed. Also, attention must be paid to the interfaces
between the components. For example, fine soil from one layer should not migrate into coarse
soil in an adjacent layer (a separation or filter layer should be used if particle migration is a
concern). In addition, adjacent materials sometimes have low shear strength along their interface
(e.g., GN/GM, GM/CCL). Thus, the design of a multi-component cover system involves careful
analysis of each component, consideration of how the components interact in a system, and
evaluation of interfaces.
The functions, materials, and design principles for the six typical cover system components of
hydraulic barrier cover systems are discussed in this chapter. Where components interact with
one another, those interactions are discussed as well. Examples of cover systems for different
applications are given at the end of the chapter.
2.2 Surface Layer
The primary functions of the surface layer are to resist erosion by water and wind, support easy
maintenance, and provide a growing medium for vegetation, if present. The surface layer can
also serve other purposes, such as promoting ET or meeting aesthetic, ecological, and site end
use criteria.
2.2.1 General Issues
Perhaps the most important concern with respect to the surface layer is the potential for erosion.
Excessive erosion can lead to exposure of underlying layers and can cause the cover system to be
ineffective. Erosion can be controlled by managing surface-water runoff (see Section 2.2.4),
minimizing seepage forces within the cover system soils (see Section 2.4), and selecting a
surface layer material that can withstand the anticipated erosive stresses due to water and wind
(see Sections 2.2.2.2 and 2.2.5).
DRAFT - DO NOT CITE OR QUOTE 2-1
-------
2.2.2 Elements of Design
Important questions that typically need to be addressed when considering the design of the
surface layer include:
• What materials are available to construct the surface layer?
• What thickness of surface layer material is needed?
• What maximum slope inclination can be used with the surface layer material while
providing acceptable erosion rates?
• For vegetated cover systems, what plant species should be established?
• How should surface-water runoff be managed?
• What minimum slope inclination is required to promote runoff after accounting for
settlement?
• What temporary and permanent erosion control measures should be used?
• How should the surface layer be constructed?
• What type and frequency of maintenance should be employed?
• What type and frequency of monitoring should be employed?
2.2.2.1 Slope Inclination
Slope inclination can be expressed in different ways, as shown in Figure 2-1. The ratio of
horizontal and vertical (e.g., 3H:1V) is perhaps the most common way of expressing the
inclination of landfill sideslopes. Slope inclinations are often expressed as a percentage when
referring to landfill top decks, runoff, or internal drainage issues. When slope stability is
analyzed, the inclination is typically expressed in degrees.
As shown in Figure 2-2, some cover systems have a relatively flat top deck and steeper
sideslopes. In such situations, the cover system components might be different in the flatter and
steeper areas. For example, the surface layer might be topsoil on the top deck and rock riprap on
the sideslopes. However, in most instances, the same components are used on both the flatter
and steeper areas.
Most landfill cover system top decks are designed to have a minimum inclination of 2 to 5%,
after accounting for settlement, to promote runoff of surface water. Slopes flatter than 2% may
allow water to pond on the surface, if localized settlements occur, and are usually avoided.
However, in some cases involving the closure or remediation of existing landfills, waste piles, or
source areas, flatter slopes may already exist and the cost to increase the slope inclination by fill
placement or waste excavation may be significant. In these cases, slightly flatter inclinations can
be considered if the future settlement potential can be demonstrated to be small, if concerns
about localized subsidence can be adequately addressed, and if monitoring and maintenance
provisions exist to repair areas of grade reversal or subsidence.
The potential for excessive erosion or slope instability increases as the cover system inclination
increases. Sideslope inclinations can range from flatter than 5H: IV to steeper than 2H: 1V.
DRAFT - DO NOT CITE OR QUOTE 2-2
-------
(a)
Percent Slope = (V/H) x 100%
Slope Angle = p = tan-1(V/H)
Ratio of Horizontal and Vertical = H:V
(b)
Figure 2-1. Slope Inclination: (a) Definitions; and (b) Example.
Relatively Flat_
Cover System
Figure 2-2. Relatively Steep and Flat Sections on a Typical Landfill Cover System.
DRAFT - DO NOT CITE OR QUOTE
2O
-3
-------
Flatter sideslope inclinations are typically associated with surface impoundments, HW landfills,
low-level radioactive waste landfills, and sites with soft remediation wastes. Some landfill cover
system sideslopes are as steep as 2H: IV (and even steeper for some old landfills). Most modern
MSW landfills have maximum cover system inclinations in the range of 4H: IV to 3H: IV, values
selected to balance the need for facility capacity with considerations related to facility
operational efficiency, waste mass and cover system slope stability, and surface erosion. Slopes
with inclinations near the flatter end of this range (4H: IV or flatter) are typically used for cover
systems when less maintenance will be performed or for projects in which erosion or slope
stability is a particularly critical issue.
2.2.2.2 Materials
In humid climates, a vegetated topsoil layer substantially reduces the potential for surface
erosion in comparison to bare ground. Vegetation serves to reduce the quantity and velocity of
runoff, reduce soil mobilization due to raindrop impact, and bind soil particles together through
root systems. Vegetation also promotes ET of infiltrating water. Alternatives to a topsoil
surface layer are typically only considered when it is difficult to maintain vegetation (e.g., on
steep slopes or in arid or semi-arid areas). At sites with this condition, the vegetative cover may
not have sufficient density to provide adequate erosion protection. Grasses and shrubs may tend
to be clumped, leaving a substantial percentage of the surface devoid of vegetation and
unprotected from wind and runoff. In such circumstances, alternative, erosion-resistant materials
may be warranted to help encourage native vegetation establishment and growth and to reduce
erosion. In this type of environment, the addition of organic matter and plant nutrients to the
surface soils and the use of soil-gravel mixtures (see Section 2.2.2.2.3), gravel (see Section
2.2.2.2.4), riprap (see Section 2.2.2.2.5), geosynthetic erosion control materials (see Section
2.2.5.4), or other materials may be required. Alternatives to a topsoil surface layer may also be
considered to achieve a desired end use for the property, e.g., a parking lot or building.
2.2.2.2.1 Topsoil
The most common material used to construct the surface layer is locally available topsoil.
Because the soils and rocks of different regions are variable, topsoils are variable, as well.
However, all topsoils tend to be relatively rich in organic matter and contain a broad mixture of
particle sizes. General information on the surface soils for a particular area of the U.S. is
summarized in the U.S. Department of Agriculture (USD A) National Resources Conservation
Service (NRCS) soils surveys. Soil surveys may be obtained from the State or local office of the
NRCS. Some of these surveys are also available online at
http ://www. statlab.iastate.edu/soils/nssc/.
Soils used for cover systems are typically classified using either engineering or agricultural soil
classification systems. The agricultural system, employed by the USDA and summarized in
Figure 2-3, classifies soil based on the relative amounts of sand, silt, and clay. A mixture of
sand, silt, and clay is called "loam." Soils that promote and sustain plant growth are typically
loamy soils. The sand in the loam provides a stable matrix that does not tend to shrink and crack
when the soil dries, and the sand helps promote good drainage. A fine material (silt and clay)
fraction is important in topsoil for retention of moisture. For these reasons, a loamy soil that
contains organic matter and nutrients is ideal for topsoil.
DRAFT - DO NOT CITE OR QUOTE 2-4
-------
30/ SANDY >
CLAY /
\ /LOAM /
Percent Sand
Figure 2-3. USDA Soil Classification System. USDA Particle Sizes: Sand, 0.05 - 2 mm;
Silt, 0.002 - 0.05 mm; and Clay, < 0.002 mm.
The design engineer should consult local agricultural specialists when evaluating the soil
proposed for the surface layer. The most appropriate type of soil to use may depend on the type
of vegetation that will be planted. Site-specific factors, such as soil pH and salinity, may be very
important.
2.2.2.2.2 Amended Topsoil
It is important that topsoil contain adequate organic matter and plant nutrients. If not,
supplements (e.g., compost, fertilizers) may be added. An increasingly common practice is to
amend topsoil with organic matter that would otherwise constitute a waste material, such as
wastewater treatment sludge or fibrous waste from production of paper. The organic matter in
these materials helps to promote growth of vegetation, and the use of these materials in surface
layers leads to productive use of a material that would otherwise be a waste material. Care
should be taken if these types of waste materials are used to ensure that surface-water runoff
from the amended topsoil is safe when discharged to surface waters. The organic amendment
DRAFT - DO NOT CITE OR QUOTE
2-5
-------
should also be demonstrated to be non-pathogenic and to not create a nuisance (e.g., odor,
vectors, etc.)
2.2.2.2.3 Soil-Gravel Mixture
At sites where excessive erosion may occur with topsoil alone, a soil-gravel mixture may be
suitable. Erosion (Ligotke, 1994) and water balance studies (Waugh, 1994) suggest that
moderate amounts of gravel (e.g., 25% by weight) mixed into topsoil can control both water and
wind erosion with little effect on the vegetation or the soil water balance. As wind and water
pass over the cover surface, some winnowing of fines from the gravel-soil mixture is expected,
creating a vegetated erosion-resistant surface sometimes referred to as a "desert pavement". The
size of gravel used in the mixture is typically in the range of 10 to 50 mm in diameter.
This design was utilized in an alternative cover system as part of a landfill research project in
Albuquerque, New Mexico. The surface treatment consisted of mixing 25% by weight pea
gravel with topsoil in the uppermost 6 inches of the fine layer of a capillary barrier. Results have
shown this to be very effective to date. (Dwyer 2001)
As another example, a 1-m thick silt loam-pea gravel mixture was used as the top deck surface
layer for a prototype cover system constructed over a contamination source area at the U.S.
Department of Energy (DOE) Hanford Site. The prototype cover system was constructed in
1994 and its performance was monitored for four years as part of a treatability study
(http://hanfordbarriers.pnl. gov/sum tests. asp). Results of the study demonstrated that the cover
system performance criteria were met or exceeded, and the cover system design components are
highly effective for the Hanford Site.
2.2.2.2.4 Gravel Veneer
A thin surface layer consisting of 10 to 50-mm diameter gravel may be used to provide more
erosion protection than a topsoil surface layer and can also result in the establishment of
vegetation. The gravel can trap seeds until they germinate. In addition, there is more near
surface moisture available for plants since there is generally less surface evaporation from a
gravel layer than from a topsoil layer. At the low matric potentials typically experienced in the
semi-arid and arid climates where a gravel surface layer may be used, finer-grained soils
generally have a higher hydraulic conductivity and, thus, higher evaporation rate than coarser-
grained soils. Consequently, after the gravel dries, the finer-grained soil below the gravel will
tend to remain moist because the overlying coarser-grained gravel layer is, at this point,
essentially non-conductive. The tendency of granular material to behave in this manner is
utilized by gardeners who apply mulch to bare soil. The mulch allows water to percolate down
to the underlying soil but shields the soil from evaporative loss of water (Kemper et al., 1994).
In comparison to a soil-gravel surface layer, a gravel veneer surface layer affects the soil water
balance. The significance of this effect has not been well studied, but its potential impact must
be acknowledged when use of a gravel surface layer is considered. The use of a gravel surface
layer reduces evaporation. However, the added vegetation established on the gravel layer and
the additional available moisture in the surface soils increases transpiration. Depending on the
site conditions, the reduction in evaporation may or may not be balanced by the increase in
transpiration.
DRAFT - DO NOT CITE OR QUOTE 2-6
-------
A gravel veneer surface layer was utilized in an alternative cover system as part of a landfill
cover research project in Albuquerque, New Mexico. The treatment installed gravel 0.6 cm
diameter and 2 to 4 cm in depth. Results have shown this method to be effective. (Dwyer 2001)
Sources of clean gravel are often limited, which means that the gravel must frequently be
quarried from rock. Before gravel is selected for the surface layer, the cost of the material
should be established to ensure that the use of gravel is practical.
When gravel is used for the surface layer, a separation layer (e.g., GT) may be necessary
between the gravel and the underlying material to prevent this latter material from being eroded
by water.
2.2.2.2.5 Riprap
At sites where is it difficult to establish and maintain vegetation, a riprap (cobble) surface layer
may be preferred. Clean riprap may adversely impact the water balance of the cover system.
Precipitation that falls on the riprap percolates downward with virtually no impedance.
Evaporation is limited because riprap has large openings and water falling though the riprap and
into the underlying soil will not be brought back by capillarity to the riprap surface for
evaporation. In addition, plants, other than occasional deep rooted plants such as shrubs and
trees, do not normally grow through the riprap and, therefore, do not remove water from the
subsoil and transpire it back to the atmosphere. Thus riprap serves as a one-way conduit for
water movement by allowing water to percolate downward into the underlying materials but
contributing almost nothing to upward water migration via ET. For example, field experiments
at Hanford, Washington, demonstrated that the placement of an unvegetated gravel surface layer
over a silty soil caused approximately half of the annual 150 mm of rainfall to percolate through
the upper 2 m of soil (Gee et al., 1992). In contrast, when silt (even without vegetation) was
exposed at the surface and not covered with gravel, there was zero percolation through the 2-m
thick soil profile during the monitoring period.
There are instances in which it may desirable to have a relatively large amount of infiltration
penetrating into the cover system. One such case involves a soil cover system constructed over
radioactive wastes that emit radon gas. For this case, surface emissions of radon can be
controlled by covering the waste with a thick, wet layer of clayey soil. Wet, clayey soils are
practically impermeable to gas. Maintaining a high water content in the soil is desirable in such
situations, and a layer of riprap at the surface can help to keep the underlying soil wet. The
increased infiltration may, however, result in increased percolation through the cover system,
and it may be more advantageous to incorporate a gas collection layer and overlying GM barrier
into the cover system.
In earthwork projects, riprap is often the most expensive material used on the project. This is
because sources of clean cobbles are fairly rare, which means that the riprap must often be
quarried from rock. Frequently, the closest source of riprap may be tens or hundreds of
kilometers from the project site. Thus, before riprap is selected for the surface layer, the cost of
the material should be established to ensure that the use of riprap is practical.
DRAFT - DO NOT CITE OR QUOTE 2-7
-------
For cover systems, riprap is often sized based upon experience, judgment, and the size of
material that is available. The typical minimum particle size of stones in riprap used for cover
systems is 10 to 300 mm. However, the minimum particle size depends on the steepness of the
slope and the anticipated water flow velocity. If relatively steep slopes are used, large, angular
stones may be necessary to maintain the stability of the stones on the slope. Some cover systems
at large landfills have somewhat irregular surfaces with high and low areas. Natural drainage
swales or channels may exist. There is more potential for higher-velocity water flow in these
swales or channels, compared to other areas, and larger stones (up to approximately 150 to 300
mm or greater) may be appropriate in such areas.
When riprap is used for the surface layer, a bedding layer (e.g., cobbles) or a separation layer
(e.g., GT) may be necessary between the riprap and the underlying material to prevent this latter
material from being eroded by water. When riprap is used to line drainage swales or channels on
the cover system, the riprap is sometimes placed on a piece of GM to limit infiltration into the
underlying cover system components. If this detail is used, an outlet should be designed to
accommodate the water collected on the GM.
As an example, a basalt riprap (less than 250 mm diameter) surface layer was used on the 2H: IV
sideslopes along the perimeter of the prototype cover system constructed over a contamination
source area at the DOE Hanford Site (http: //hanf ordb arri er s. pnl. gov/sum si ope. asp). As
previously mentioned in Section 2.2.2.2.3, the performance of the prototype cover system was
monitored for four years and found to be satisfactory.
2.2.2.2.6 Asphaltic Concrete
Asphaltic concrete is a mixture of aggregate (usually sand and gravel) and asphalt, sometimes
with additional materials such as polymers. Heated asphalt is mixed with aggregate, spread in a
thin layer (typically 50 to 100 mm thick), and compacted with heavy, steel vibratory drum
rollers. Asphaltic concrete can be placed as a single layer or in multiple layers.
Asphaltic concrete can be quite permeable unless special attention is given to minimizing air
voids during mixing and application (Repa et al., 1987). To achieve low hydraulic conductivity,
1.5 to 2 times more asphalt is used than is typical for roadway pavements. This type of asphaltic
concrete is referred to as "low-permeability asphaltic concrete." Both ordinary and low-
permeability asphaltic concrete have been used in cover systems. In some cases, the low-
permeability asphaltic concrete layer is the only cover system component and functions as the
surface layer and hydraulic barrier.
A low-permeability asphaltic concrete layer should not be considered as a permanent hydraulic
barrier, unless it is maintained. Asphalt becomes brittle over time as a result of exposure to
ultraviolet radiation and oxygen. In addition, an asphaltic concrete layer in a cover system may
develop cracks due to differential settlement of underlying waste. If the intent is to maximize
design life, the asphaltic concrete layer should normally be buried beneath a protection layer and
not subjected to differential settlements that would induce cracking.
The following are examples of cover systems in which asphaltic concrete was used as the surface
layer. One case involved a 1-ha area of contaminated soil that was located next to an office
DRAFT - DO NOT CITE OR QUOTE
-------
building. The cover system was paved with ordinary asphaltic concrete and used as a parking
lot. In a second case, a small section of a landfill cover system was paved with low-permeability
asphaltic concrete to create an area that could be used to park maintenance vehicles. The third
case was a remediation project in which there was particular concern for minimizing or
eliminating erosion. Again asphaltic concrete was used as the surface layer. In the latter two
cases, the asphaltic concrete was a low-permeability material that contained an asphalt
application rate intended to produce a hydraulic conductivity of 1 x 10"9 m/s or less. In both of
these cases, the asphaltic concrete served as a surface layer and hydraulic barrier.
The National Risk Management Research Laboratory of the EPA is currently evaluating the
application of a low-permeability asphaltic concrete cover system to two CERCLA sites under
the Superfund Innovative Technology Evaluation (SITE) Program (http://www.epa.gov/ORD/
SITE/). Each cover system consists of a 100-mm thick layer of proprietary-blend low-
permeability asphaltic concrete.
2.2.2.2.7 Other Materials
Practically any material, including articulated block systems, some construction and demolition
wastes, and some lightweight manufactured aggregates (e.g., expanded shale), could potentially
be used as a material in a surface layer or could be mixed with other materials and used for the
surface layer. However, if something other than soil, gravel, or riprap is considered, it will
generally be because there is a special desire or incentive for utilizing a particular material.
Alternative materials should be considered if they are safe, stable, and can meet applicable
design criteria.
2.2.2.3 Thickness
The minimum thickness of the surface layer is established based on consideration of the rooting
depth of any surface vegetation, anticipated erosion rate, and construction tolerances. With
respect to the latter, it is usually not practical to construct a layer thinner than about 0.15 m using
typical earth moving equipment. If topsoil or a topsoil-gravel mixture is used, the soil should be
thick enough to accommodate a healthy growth of plant roots. For shallow-rooted plants such as
certain grasses, a 0.15-m thick layer of soil usually provides adequate rooting depth. Thus, the
minimum thickness of a vegetated surface layer is generally 0.15 m. If plants with deeper roots
are planted or represent a desirable climax community, the thickness of the topsoil should be
increased to accommodate root growth. The underlying protection layer (if present) may also
accommodate plant roots, in which case 0.15 m of topsoil may be all that is needed for the
surface layer.
In some instances, the surface layer and protection layer are constructed from the same type of
material, making it impossible to distinguish one layer from the other. The combined layers may
be referred to as "cover soil" or "cover material". If the surface and protection layers are
combined into a cover soil, then the minimum thickness of the cover soil should be evaluated
considering the plant rooting depth. A typical minimum thickness of the cover soil is 0.45 to 0.6
m for cover systems with hydraulic barriers. For cover systems with ET or capillary barriers,
EPA recommends a minimum cover soil thickness of 0.9 m or greater (see Section 3.2.5).
Thicknesses greater than 1 m are occasionally used to provide a suitable medium for growth of
plants in relatively arid areas, which commonly have deep-rooted plants. Greater thicknesses of
DRAFT - DO NOT CITE OR QUOTE 2-9
-------
cover soil may also be needed to provide a hydraulic barrier with protection from desiccation or
frost.
If gravel or riprap is used for the surface layer, the minimum thickness is usually 0.15 m or twice
the average particle size of the material, whichever is larger.
If asphaltic concrete is used for the surface layer, the minimum thickness should be determined
from an analysis of vehicular loading, but would typically be in the range of 75 to 150 mm.
2.2.3 Vegetation
Selection of plant species is an important consideration in the design of a vegetated surface layer.
The vegetation serves several functions:
• Plant leaves intercept some of the rain before it impacts the surface layer, thereby
reducing the energy of the water and the potential for erosion.
• Plant vegetation also helps dissipate wind energy.
• The shallow root system of plants enhances the surface layer resistance to water and wind
erosion.
• Plants promote ET of water, which increases the available water storage capacity of the
cover soils and decreases drainage from these soils.
• A well-vegetated surface layer is generally considered more natural and esthetically
pleasing than an unvegetated surface layer.
In selecting the appropriate vegetation for a site, the following general recommendations are
offered:
• Locally-adapted, low-growing (less than 1 m high) grasses and shrubs that are
herbaceous or woody perennials should be selected. Native plants are recommended to
maintain long-term ecological stability.
• The plants should survive drought and temperature extremes. They should also tolerate
inhospitable site conditions (e.g., exposure to landfill gas).
• The plants should contain roots that will penetrate deep enough to remove moisture from
beneath the surface but not so deep as to disrupt the drainage layer, hydraulic barrier, or
gas collection layer.
• The plants should be capable of thriving with minimal addition of nutrients.
• The plant population should be sufficiently diverse to provide erosion protection under a
variety of conditions.
• The plants should not be an attractant to burrowing wildlife.
• The vegetative cover should be capable of surviving and functioning with little or no
maintenance (e.g., without excessive irrigation, fertilization, and mowing).
DRAFT - DO NOT CITE OR QUOTE 2-10
-------
Guidance on selection of vegetative materials is found in Wright (1976), Thornburg (1979), Lee
et al. (1984), and EPA (1985). These references provide information about plant species,
seeding rate, time of seeding, and areas of adaptation. Growth information for a number of plant
species is available in the USDA Plant database at http://plants.usda.gov/. Local plant
specialists, such as the NRCS, are usually consulted to select the appropriate mixture of seeds for
a site. Local NRCS and Department of Transportation specifications may also be useful.
Experience also is very helpful, and once a seed mixture has been shown to provide satisfactory
performance in a particular region, it tends to continue to be used.
At many sites with cover systems located in humid and temperate parts of the country, the cover
systems are seeded with a mixture of grasses. The mixture may contain several grass species to
provide diversity in the grass population, promote vegetative growth for as much of the year as
possible, and maintain a vegetative layer with the desired mixture of shallow- and medium-depth
roots. Information on grasses is available in Hanson and Juska (1969), who subdivide the U.S.
into the four regions shown in Figure 2-4. Native or locally-adapted grasses that they generally
recommend for permanent vegetative covers are listed in Table 2-1.
LEGEND
I I Region 1: Cool, humid
I I Region 2: Warm, humid
LJ Region 3: Warm and and semi-and
[_] Region 4: Cool, arid and serm-and
Figure 2-4. Major Regions of Grass Adaptation in the U.S. (modified from Hanson and
Juska, 1969).
DRAFT - DO NOT CITE OR QUOTE
2-11
-------
Table 2-1. Grass species recommended for use as permanent vegetative covers in the
four regions of grass adaptation (modified from Hanson and Juska, 1969).
Region Species Seeding1! Seeding Comments
ime Rate2
(kg/ha)
Cool-humid
(Region 1)
Warm-humid
(Region 2)
Warm-arid &
semi-arid
(Region 3)
Cool-arid &
semi-arid
(Region 4)
Kentucky bluegrass
(Poa pratensis L.)
Tall fescue
(Festuca arundinacea
Screb.)
Perennial ryegrass
(Lolium perenne L. )
Smooth brome
(Bromus inermis Leyss.)
Redtop
(Agrostis alba L. )
Weeping lovegrass
(Eragrostis curvula Schrad.)
Bermudagrass
(Cynodon dactylon L.)
Bahiagrass
(Paspalum notatum Fluegge)
Zoysia
(Zoysiajaponica Steud)
St. Augustine grass
(Stenotaphrum secundatum
Kuntze)
Bermudagrass
(Cynodon dactylon L.)
Buffalograss
(Buchloe dactyloides Englem.)
St. Augustine grass
(Stenotaphrum secundatum
Kuntze)
Bermudagrass
(Cynodon dactylon L.)
Buffalograss
(Buchloe dactyloides
Englem.)
Sideoats grama
(Bouteloua curtipendula Torn)
Fairway wheatgrass
Spring &
Fall
Spring &
Fall
Spring &
Fall
Spring &
Fall
Spring &
Fall
Spring &
Early
Summer
Spring &
Early
Summer
Early
Summer
Summer
Early
Summer
Spring
Spring
Early
Summer
Early
Summer
Spring &
Early
Summer
Spring
Spring
20
40
40
20
15
5
10
20
See
Reference
See
Reference
10
25
See
Comment
10
25
35
25
Do not use named varieties
Use K-31 or Alta varieties; can
winter kill north of Interstate 80
Do not use named varieties
Use southern type except in
extreme northern part of region
Not very tolerant of mowing; good
for wet conditions
Use in southern % of region only
since less winter hardy than other
species
Do not use named varieties
Do not use named varieties unless
cold tolerance is important
Propagated vegetatively
Propagated vegetatively; common
is coarser textured than named
varieties
Do not use named varieties
Use only in the eastern % of the
region
Use only in extreme southern part
of region and at low elevations
Do not use named varieties; use
only in extremely southern part of
region
Use only in eastern % of region
Use Blue grama (Bouteloua
gracilia Lag.) if less than 380 mm
precipitation
Best adapted to northern 1/4 of
(Agropyron cristatum
Gaertn)
region; use Crested wheatgrass
(A. desertorum Schult.) in the
southern part at elevations of
1,500 to 2,700m
1 For species that can be seeded spring and fall, fall seedings are almost always more successful.
2 Seeding rates assume single species. Reduce rates by the number of components in mixtures.
Minimum % pure live seed of 70 is assumed (% pure live seed = % germination x purity).
If the % pure live seed is less than 70, increase seeding rate accordingly.
DRAFT - DO NOT CITE OR QUOTE
2-12
-------
Sometimes the vegetation is selected to maximize ET of water. For example, O'Donnell et al.
(1997) describe the use of juniper plants to minimize infiltration of water through a cover
system. Hybrid poplar trees, planted at a high density (e.g., 2,700 trees/ha), have also been used
for the same application (Licht et al., 2001).
For cover systems in humid or temperate climates vegetated with grasses, the grasses are usually
mowed periodically to discourage the growth of shrubs, trees, or other types of deep-rooted
plants. Deep-rooted plants are usually undesirable because their root systems could plug the
drainage layer or penetrate the hydraulic barrier, if it consists of only a CCL or GCL without an
overlying GM. Trees can also create problems if they are blown over, uprooting large masses of
soil and leaving a crater in the surface.
For sites designed to allow the development of climax communities, plant roots are typically
deeper than for sites vegetated only with grasses. To prevent clogging of the drainage layer by
plant roots, the thickness of the cover soils is increased or the drainage layer is sometimes treated
with a biocide. Alternatively, the cover system is designed with relatively shallow sideslopes so
that the ability of the drainage layer to function is not as critical. For example, native plants,
including coastal sagebrush, were established on several closed landfills with thick ET cover
systems in southern California in the late 1990's. When the native plants on these covers were
studied to assess their growth characteristic, the roots of some of the native species had
penetrated up to 2 m into the cover system soils.
To help in the initial establishment of vegetation, adequate soil nutrients should be available. In
addition, soils detrimental to vegetation growth (e.g., soils with high salt contents) should be
avoided. While soil amendments will improve the soil's characteristics as a rooting medium, any
additional processing or amendments will increase costs.
2.2.4 Surface-Water Control
Surface-water runoff from the cover system should be controlled using a surface drainage
system. The channelization of runoff is critical with respect to managing flow and controlling
erosion. The drainage system may consist of a network of swales, ditches, downchutes, drop
pipes, and culverts. Each component of the drainage system should be designed for the peak
flow conditions anticipated from the design storm. Downchutes represent a particular challenge
due to the high water velocities that occur on steep slopes. Flows from the cover system are
typically directed to sediment traps, basins, and/or ponds to minimize the release of sediments
and control rates of water flow from the site.
The design of a surface drainage system often constitutes a significant exercise in surface-water
hydrology. The process can be very complex, involving statistical analysis of storm events,
prediction of runoff for situations where minimal quantitative data exist, consideration of the
potential occurrence of storms during interim stages of landfill development, consideration of
changing cover system inclinations over time as the underlying waste settles, and other
complications.
DRAFT - DO NOT CITE OR QUOTE 2-13
-------
It is common practice to construct swales and ditches on cover systems with long vegetated
sideslopes to intercept runoff and water from any cover system drainage layer outlets (Figure 2-
5). Swales may be formed by constructing soil add-on berms on a uniformly sloping cover
system (Figure 2-5(a)) or by constructing benches into the cover system sideslopes (Figure 2-
5(b)). Ditches may be constructed adjacent to cover system access roads (Figure 2-5(c)). The
swales and ditches are often connected to armored downchutes or to drop pipes, which convey
runoff from the cover system sideslopes. A supplemental hydraulic barrier may be installed
beneath the surface layer of swales, ditches, and downchutes to decrease the potential for
infiltration of water into underlying cover system components. If the cover system surface layer
consists of riprap or asphaltic concrete, surface drainage features, such as swales and ditches,
may not be necessary.
The vertical spacing of swales and ditches on a cover system slope should be designed
considering the need to manage surface water and limit erosion. In many cases, the spacing is
controlled by erosion concerns (see Eq. 2-5 in Section 2.2.5.4 and Eq. 2-9 in Section 2.2.5.5.3)
and is a function of slope inclination, surface layer material and vegetation properties, rainfall
intensity, and other factors. As a general rule of thumb, surface-water interception may be
necessary on cover system sideslopes at intervals of 10 m vertically or 30 m along the slope,
whichever produces more frequent benches. Leaving out benches altogether on slopes with
lengths greater than approximately 30 to 50 m may lead to excessive erosion and is usually
avoided for slopes with inclinations greater than 5%. Erosion rills forming gullies as deep as 1 m
can develop, and hundreds of cubic meters of soil can be washed away in a few days of
inclement weather if adequate surface water controls are not employed. The actual vertical
spacing of swales and ditches on a cover system should be based on local factors and detailed
hydraulic and erosion analyses and should not be arbitrarily established.
Since swales, ditches, and downchutes convey concentrated flow from cover systems, they may
need to be armored with turf reinforcement mat, riprap, or other material (see Section 2.2.5.7) to
have adequate resistance to erosion. Extra erosion control measures may also be required at
surface drainage system transitions (e.g., at the intersection of a swale and a downchute or down
pipe).
Surface drainage system design typically involves the following general steps: (i) divide the
cover system into several distinct drainage areas, as necessary; (ii) estimate the hydrologic
properties of each area using size, soil type, and vegetative cover type; (iii) evaluate the rate of
runoff from the design storm for each drainage area and the peak rate of runoff at each surface
drainage system component; and (iv) size each component of the surface drainage system to
handle the estimated peak flow associated with it. When the drainage system includes a
sedimentation pond for stormwater management, the required storage volume of the pond also
needs to be evaluated.
The design storm is usually specified for temporary and permanent conditions in federal, state,
and local waste management, flood control, and soil conservation regulations. For example,
federal regulations for MSW landfills (40 CFR §258.26) and HW landfills (40 CFR §264.301(h)
and 40 CFR §265.30l(h)) require these facilities to be designed to manage at least the 24-hour
storm with a 25-yr return period. For containment applications with a higher level of risk to
DRAFT - DO NOT CITE OR QUOTE 2-14
-------
TYPICAL COVER SYSTEM
Cover Soil
Drainage Layer
GM
GCL
Soil Foundation Layer-
Figure 2-5. Details of Typical Swales and Ditches for Cover Systems (from Koerner and
Daniel, 1997): (a) Swale Constructed with Add-on Berm; (b) Swale
Constructed by Benching Sideslopes; and (c) Ditch Sometimes Constructed
Adjacent to Access Road.
DRAFT - DO NOT CITE OR QUOTE
2-15
-------
human health and the environment, such as for low-level radioactive waste disposal facilities, the
design storm may be developed based on human health risk, statistical analysis of precipitation
events, the PMP event, and other factors. As an example, the 2,000-yr design storm was
considered when designing the on-site disposal facility at the DOE Fernald Environmental
Management Project site.
Several urban drainage models are available for surface-water analysis for small (i.e., less than
about 500 ha) urban watersheds. Two of the most commonly used models are: (i) the "rational
method"; and (ii) the USDA Soil Conservation Service (SCS) Technical Release Number 55
(TR-55) method. (Note that the SCS is now the NRCS.) Both of these methods are described
below.
The "rational method" is one of the simplest and best-known analysis methods routinely applied
in urban hydrology. It is commonly used in civil engineering applications and is a method
approved by the DOE (1989) for design of cover systems for sites regulated by the Uranium Mill
Tailings Radiation Control Act (UMTRCA) of 1978 (i.e., Uranium Mill Tailings Remedial
Action (UMTRA) sites). The rational method is based on the assumption that rainfall occurs
uniformly over the watershed and at a constant intensity for a duration equal to the time of
concentration. This method is typically used for areas under 40 ha in size. Using the rational
method, the peak rate of runoff, q (m3/s/m), is calculated as:
q = cirAbF (Eq. 2.1)
where: c = runoff coefficient (dimensionless) and is equal to runoff divided by precipitation, ir =
rainfall intensity (m/s) for the period of interest; Ab = area of the drainage basin or subbasin per
basin or subbasin width (m2/m); and F = flow concentration factor (dimensionless).
Input values to the rational equation are as follows:
• The runoff coefficient is a function of ground cover, soil antecedent moisture, ground
slope, and other factors. Runoff coefficient values are given in many hydrology
textbooks and can range from near zero for shallow-sloping, grassed sandy soils to
essentially 1.0 for impervious cover. Typical runoff coefficient values for different
vegetation and slope conditions are shown in Table 2-2. For storms with return periods
longer than 100 years, DOE recommends the use of c = 1.0 (DOE, 1989).
• Rainfall intensity is calculated as:
ir = d / tc (Eq. 2.2)
where: d = depth of rainfall in time of concentration from a storm with a certain return
period (m); and tc = time of concentration (s). The equation used to calculate the time of
concentration depends on the surface layer material. For a soil, vegetated, or paved
surface layer, the time of concentration can be calculated using the method of Brant and
Oberman presented in DOE (1989):
DRAFT - DO NOT CITE OR QUOTE 2-16
-------
t = 0.0328 C.
1/3
s(02
(Eq. 2.3)
where: Cs = surface layer coefficient (dimensionless) and is 0.5 for paved areas, 1.0 for
unvegetated soil; and 2.5 for turf; Lf = length of overland flow path (m); S = slope
inclination (dimensionless); and all other terms are as defined previously. For a riprap
surface layer, the time of concentration can be calculated using the method of Kirpich
presented in U.S. Nuclear Regulatory Commission (NRC) (1990):
=0.0192
(Lf)3
(Eq. 2.4)
where: Hf = elevation difference along flow path (m), and all other terms are as defined
previously. Whatever the surface layer, DOE (1989) recommends that the minimum time
of concentration used in Eq. 2.2 be no less than 150 seconds. This is because for very
small values of tc, small decreases in tc will cause relatively large increases in ir, resulting
in over-conservative estimations of the peak rate of runoff. Values for d in Eq. 2.2 are
obtained from rainfall intensity maps (e.g., Hershfield, 1961; Miller et al., 1973; Hansen
etal., 1982).
The flow concentration factor accounts for flow possibly concentrating in rills and
gullies. When calculating the peak rate of runoff to size drainage structures, F = 1.
When evaluating the potential for gully formation (see Section 2.2.5.5), the flow
concentration factor generally ranges between 1 and 3. For vegetative covers, Caldwell
and Reith (1993) recommend using flow concentration factor values between 2 and 3.
For riprap-lined channels, Abt et al. (1987, 1988) recommend using values between 1 and
3.
Table 2-2. Runoff coefficient values (modified from Barfield et al., 1983).
Vegetation and
Slope Conditions
Woodland
Flat, 0-5% slope
Rolling, 5-10% slope
Hilly, 10-30% slope
Pasture
Flat, 0-5% slope
Rolling, 5-10% slope
Hilly, 10-30% slope
Cultivated
Flat, 0-5% slope
Rolling, 5-10% slope
Hilly, 10-30% slope
Soil Texture
Open sandy
loam
0.10
0.25
0.30
0.10
0.16
0.22
0.30
0.40
0.52
Clay and silty
loam
0.30
0.35
0.50
0.30
0.36
0.42
0.50
0.60
0.72
Tight clay
0.40
0.50
0.60
0.40
0.55
0.60
0.60
0.70
0.82
DRAFT - DO NOT CITE OR QUOTE
2-17
-------
The TR-55 method (SCS, 1986a) is based on the unit hydrograph method of analysis, and, thus,
unlike the rational method, it can be used to calculate runoff volume and sediment pond storage
volume as well as the peak rate of runoff. It also can better accommodate sites with varying
topography and surface layer characteristics. Like the rational method, TR-55 starts with a
"runoff coefficient", called a "runoff curve number"(CN) in TR-55, and a rainfall amount
uniformly imposed on a watershed over a specified time. At the start of a precipitation event,
some rainfall is considered lost to plant interception, evaporation, infiltration into the surface
soil, and storage in surface depressions. After the initial loss, called the "initial abstraction" is
satisfied, any additional rainfall may generate runoff. TR-55 calculates the runoff volume
considering the initial abstraction and then transforms the runoff into a hydrograph using unit
hydrograph theory and routing procedures that depend on runoff travel time through each
segment of the watershed. Four different unit hydrographs are used to represent storm events
across the U.S. Two of the rainfall distributions, Types IA and I, are representative of the Pacific
maritime climate that occurs in Alaska, the western half of Washington and Oregon, and most of
California. The Type 3 distribution is representative of the Atlantic and Gulf of Mexico coastal
areas. The Type 2 distribution is similar to the Type 3 and occurs in the rest of the country.
After hydrographs for watershed segments have been routed to a specific location, the peak
runoff rate at that location can be calculated by adding the hydrographs.
Once the design flow rate is determined, the surface drainage system can then be designed to
handle the flow. Open channel flow in swales, ditches, or downchutes is analyzed for the depth
and velocity of water to ensure that the system has adequate capacity to convey flow with
sufficient freeboard and that flow velocities are not greater than those specified for the specific
drainage structures. The book by Chow (1959) is often used as a reference for analyzing open
channel flow. Down pipes can usually be designed using open channel flow equations. Standard
equations for flow in pipes are presented in numerous fluid hydraulics texts and provided by the
pipe manufacturers.
2.2.5 Erosion Protection
2.2.5.1 Overview
Excessive erosion of the surface layer has been a significant problem for a number of cover
systems. Gullies extending to a depth of 100 to 200 mm are not unusual. In the extreme, the
underlying drainage and barrier layers can be eroded. Although erosion problems can often be
addressed as a maintenance activity, there have been instances of major erosion that displaced
hundreds of cubic meters of soil from inadequately protected landfill covers. Swope (1975)
studied 24 landfill cover systems in the U.S. and found that 33% had slight erosion, 40% had
moderate erosion, and more than 20% had severe erosion. Johnson and Urie (1985) report that
erosion can be made more severe by the installation of a hydraulic barrier within a landfill cover
system. Without an overlying drainage layer, the barrier can cause the cover soils to become
soaked. Saturation decreases soil strength, increases particle detachment, and increases erosion
potential (NRCS, 1998a). Even in natural soil systems, cover soils over a compacted layer on a
steep slope may slide downslope as a mass if the soils become saturated (NRCS, 1998a).
Gross et al. (2002) described several cases of significant cover system erosion, including one for
a cover system with 60-m long 3H: IV sideslopes (see Section 7.6.2). This cover system
included sand berms to divert surface-water runoff from the top deck of the landfill to riprap-
DRAFT - DO NOT CITE OR QUOTE 2-18
-------
lined downchutes on the landfill sideslopes. Sand add-on berms were also located at a few
locations on the sideslopes. The sand berms on the top deck developed gullies at several
locations allowing concentrated flow of runoff down the sideslopes. Though this cover system
included a sand drainage layer, it was not designed to outlet on the cover system and did not
have sufficient capacity to convey drainage from the cover system top deck and sideslopes. The
combination of inadequate management of surface water, insufficient drainage layer capacity,
and long steep sideslopes contributed to the erosion problems at the site (Figure 2-6).
2.2.5.2 Nature of Erosion
Soil erosion involves a process of both particle detachment and transport by water or wind. It is
initiated by drag, impact, or tractive forces acting on individual particles of soil at the surface.
Water erosion starts when raindrops impact soil particles, dislodging them and sending them
• *
Figure 2-6. Deep Gullies Through the Topsoil and Sand Drainage Layers Exposed the
GM Barrier on 60-m Long, 31-1:1 V Landfill Sideslopes.
upward into the air and some distance away. As water collects on the soil surface, it begins to
run off in small rivulets and then sheets of uniform flow. The sheet flows carry soil particles
dislodged during impact and particles dislodged by tractive forces exerted from the flow. As the
sheet flows move downslope, the flows concentrate due to irregularities in the soil surface and
topography. The resulting concentrated flows cut more deeply into the surface, creating small
channels called rills that may be tens of millimeters deep. Rill erosion accelerates with increase
in runoff, slope inclination, and slope length. Rills can be removed from a slope and will return
in different patterns and shapes. If rill development is allowed to progress, the rills will form
DRAFT - DO NOT CITE OR QUOTE
2-19
-------
deep cuts in the soil surface and become gullies. Because of the high velocities of flow in
gullies, massive removal of soil is possible. Gullies may be several feet or more deep and, unlike
rills, can generally not be repaired with a simple tilling of the soil surface. They also grow and
deepen, as sheet flow passing above the headcut of a gully exerts forces on the flow channel
boundary and removes accumulated soil debris from the channel. The types of water erosion that
may occur on a cover system are illustrated in Figure 2-7.
The erosion potential of soil is primarily a function of the size of the soil particles, interparticle
cohesive forces, and the velocity of the transporting fluid (air or water). This relationship is
illustrated in Figure 2-8. Erosion potential increases with decreasing particle size and increasing
velocity of the transporting fluid. Clays, however, which have the smallest particle size, also
have cohesion, meaning that they stick to each other, which helps to prevent erosion. Some
sodium-rich clays do not adhere to one another very well and, therefore, are highly vulnerable to
erosion. Such clay soils are called "dispersive clays." Several tests exist to identify potentially
dispersive clays (Sherard et al., 1976). Silt has a small particle size but lacks cohesion. Silt is,
therefore, almost always highly erodible. Neither dispersive clays nor silts should be used for
the surface layer, unless it can be clearly demonstrated that erosion will not be a problem.
In arid and semi-arid climates, which have sparse vegetation and dry surficial sediments, winds
can cause significant erosion. Winds can pick up and carry in suspension the lighter, less dense
soil constituents (e.g., organic matter, clays, and silts with particles sizes primarily less than 0.1
mm) (Gray and Sotir, 1996). This is why soil-gravel mixtures or gravel veneers are often
considered as a surface layer for cover systems constructed at arid and semi-arid sites. By
transporting the lighter soil particles, wind removes the most fertile part of the soil and lowers
soil productivity (Lyles, 1975). The majority (approximately 62 to 97%) of wind-eroded soil is
carried near the ground surface at heights less than 1 m. Windbreaks can be used to impede soil
movement within this height interval. Though wind can cause significant soil loss, most erosion
of soil covers in arid and semi-arid areas is caused by water.
Raindrop Erosion
Sheet Erosion
Rill Erosion
Gully Erosion
Figure 2-7. Types of Water Erosion That May Occur on a Cover System.
DRAFT - DO NOT CITE OR QUOTE 2-20
-------
70
60
50
•a 40
03
.si
;> 30
'o
20
10
Erosion
(Detachment)
Deposition
0.1
0.2
0.4
1.0
2.0
Size of Particles (mm)
Figure 2-8. Relationship Between Erosion Mechanism (Air or Water), Particle Size and
Fluid Velocity (Garrels, 1951 as referenced by Mitchell, 1993).
2.2.5.3 Short-Term and Long-Term Erosion
The cover system design should address the potential for short-term erosion (i.e., before a good
stand of vegetation is established), and make use of temporary erosion-control measures as
necessary. The design should also address long-term erosion after vegetation has been
established especially for the site-specific rainfall or wind event. Erosion can be damaging not
only to the cover system but also to areas into which eroded soil is deposited. It is also important
that constructed erosion-control measures be installed correctly and maintained.
The timing for completion of cover system construction can impact the potential for erosion. In
northern climates, the end of the construction season coincides with the end of the growing
season. A common problem is that the cover system is seeded at a time of year that is not
conducive to growing grass. In some climates, it may be impossible to initiate growth of the
vegetative cover during certain parts of the year. It is recommended that construction be
scheduled to allow vegetation to become established as soon as practicable and before the end of
DRAFT - DO NOT CITE OR QUOTE
2-21
-------
the growing season, if at all possible (EPA, 2002). If this is not achievable, erosion control
materials may be needed to protect the surface layer.
The construction contractor is usually made responsible for maintaining temporary erosion
control measures and repairing erosion damage during and shortly after construction. However,
the contractor usually has only limited expertise in soil erosion control. Further, the contractor is
not privy to the design decisions that affect the potential for severe short-term erosion. Thus,
caution should be exercised in placing responsibility upon the contractor, who may be ill
equipped to make informed decisions about appropriate erosion control measures. It is
recommended that the design engineer consider carefully the potential for and consequences of
short-term erosion and be proactive in specifying appropriate control measures (e.g., silt fences,
rolled erosion control materials, sediment traps, hay bales, etc.) in the construction documents.
The NRCS has developed conservation practice standards for a number of erosion control
measures. Most state NRCS offices have websites with downloadable conservation practice
standards. There may also be local requirements and standards for erosion control.
The NRCS (2000) makes the following recommendations to limit short-term erosion during
construction:
• cover disturbed soils as soon as possible with vegetation or other materials (mulch) to
reduce erosion potential;
• divert water from disturbed areas;
• control concentrated flow and runoff to reduce the volume and velocity of water and
prevent formation of rills and gullies;
• minimize the length and steepness of slopes (e.g., use benches);
• prevent off-site sediment transport;
• inspect and maintain any structural control measures;
• where wind erosion is a concern, plan and install windbreaks;
• avoid soil compaction by restricting the use of trucks and heavy equipment to limited
areas; and
• break up or till soils compacted by grading prior to vegetating or placing sod.
Long-term erosion is an important consideration in the design of the surface layer. In spite of the
admittedly approximate nature of predictive equations for erosion control, most cover systems
will require an analysis of long-term and, sometimes, short-term erosion. Typical design criteria
are as follows:
• The design sheet and rill erosion rate should not be exceeded. Although it is advisable to
select allowable rates of soil erosion on a project-specific basis, many design engineers
follow the general guidance that the design sheet erosion rate not exceed 4.5
tonnes/ha/year (EPA, 1991).
• Using the sheet and rill erosion rate from this calculation, the thickness of cover soil at
the end of the design life should be calculated to verify that there is adequate thickness
DRAFT - DO NOT CITE OR QUOTE 2-22
-------
remaining and that sheet and rill erosion has not progressed through the cover soil and
into the underlying layers. There should also be sufficient soil thickness to support
vegetation and provide freeze-thaw protection of a CCL barrier, if present.
• The surface layer should resist gully formation under the tractive forces of runoff from
site-specific design storm.
• If the potential for wind erosion is a concern (e.g., for some arid sites), wind erosion
should also be evaluated.
The analysis of sheet and rill erosion, gully formation, and wind erosion is discussed in Sections
2.2.5.4, 2.2.5.5, and 2.2.5.6, respectively.
2.2.5.4 Sheet and Rill Erosion
2.2.5.4.1 Universal Soil Loss Equation
The average annual rate of soil loss by water erosion is often estimated by design engineers using
some form of USDA's Universal Soil Loss Equation (USLE). The Revised USLE (RUSLE)
(Renard et al., 1997) is an improved version of USLE and is currently recommended by the
USDA for calculation of soil loss. RUSLE was developed to estimate soil loss caused by
raindrop impact and sheet flow (collectively referred to as "interrill" erosion) plus rill erosion. It
is derived from the theory of erosion processes, data from natural rainfall plots, and results for
rainfall-simulation plots.
The RUSLE method is directed toward the prediction of erosion from construction sites, mined
lands, reclaimed lands, and other disturbed areas. The areal extent and surfacing of many cover
systems provide similar conditions to those for the above landforms. RUSLE, however, is
limited to the estimation of average annual erosion rates and cannot establish erosion from
specific events. The soil loss prediction represents an average for many storms and years. In
addition, there is no direct method within the RUSLE procedure to determine the depth or
magnitude of gully erosion on a cover system. It is, therefore, recommended that this method be
used with another method that considers gully development.
The RUSLE is expressed as:
As = Re K (LS) C Pc (Eq. 2.5)
where: As = average annual soil loss by sheet and rill erosion (tonnes/ha/yr); Rg = rainfall
energy/erosivity factor (dimensionless) and is a measure of rainfall energy and intensity rather
than just rainfall amount; K = soil erodibility factor (dimensionless), is a measure of the relative
resistance of a soil to detachment and transport by water, and varies based on seasonal
temperature and rainfall; LS = slope length and steepness factor (dimensionless) and is the ratio
of soil loss from a given field slope to that from a slope that has a horizontal length of 22.1 m
(from the origin of sheet flow to the point where runoff is concentrated in a defined channel) and
a steepness of 9%; C = vegetative cover and management factor (dimensionless) and is the ratio
of soil loss from land cropped under the specified conditions to the corresponding loss from
clean-tilled, continuous fallow; and Pc = conservation support practice factor (dimensionless) and
DRAFT - DO NOT CITE OR QUOTE 2-23
-------
is the ratio of soil loss with a specific support practice to the corresponding soil loss with uphill
and downhill tillage.
Input values for RUSLE are developed using site-specific information and the database that is
part of the RUSLE computer program. Version 2 of the program can be downloaded from
http://bioengr.ag.utk.edu/rusle2/.
Using As computed from Eq. 2.3, the thickness of cover soil at the end of the cover system
design life can be calculated to verify that there is cover soil remaining and that the thickness of
this remaining cover soil is sufficient to protect the any CCL component of the cover system.
2.2.5.4.2 Water Erosion Prediction Project (WEPP) Model
The WEPP model was developed in the 1980's when an increasing need for improved erosion
prediction technology was recognized by the major research and action agencies of the United
States Department of Agriculture and Interior, including the Agricultural Research Service
(ARS), Natural Resource Conservation Service (NRCS), Forest Service (FS), and Bureau of
Land Management (BLM). In 1985, these agencies embarked on a 10-year research and
development effort to replace the Revised Universal Soil Loss Equation. Some of the
differences between the WEPP model and the RUSLE are as follows:
• The RUSLE equation is based on undisturbed agricultural and rangeland top soil
conditions, whereas any kind of soil can be described with WEPP. Thus, WEPP is well
suited to describe a landfill cover, which is a disturbed condition.
• The WEPP model is capable of predicting erosion and deposition in more complex
situations, such as when berms are involved. WEPP can predict the erosion on a cover as
well as the deposition in berm channels in the watershed mode. The WEPP model's
ability to determine runoff and channel flow can also aid in determining stability issues
with berms, such as overtopping. RUSLE can only predict the upland erosion between
berms.
• RUSLE can only predict average annual upland erosion. WEPP's climate generator
includes stochastically generated events. This is an important point in arid environments
where there are very few precipitation events annually, but when they occur, they are
often torrential events that have major impacts on the site. Thus, a landfill in an arid
climate is unlikely to fail in an average year, whereas, it is very likely to fail in a year
when a major storm event has occurred. WEPP can predict the impacts from a major
storm event, but RUSLE cannot.
Additional information regarding the WEPP model, software, and documentation can be found
at: http://topsoil.nserl.purdue.edu/nserlweb/weppmain/wepp.html.
DRAFT - DO NOT CITE OR QUOTE 2-24
-------
2.2.5.5 Gully Erosion
2.2.5.5.1 Overview
The concentration of runoff under many circumstances encourages the formation of rills, which,
if unchecked, grow into gullies. This is arguably the most severe type of erosion of cover
systems soils at landfill and waste remediation sites.
The dynamics of gully formation are complex and not completely understood. Gully growth
patterns are cyclic, steady, or spasmodic and can result in the formation of continuous or
discontinuous channels. Gully advance rates have been obtained by periodic surveys,
measurements to steel reference stakes or concrete-filled auger holes, examination of gully
changes from small-scale maps, or from aerial photographs. Studies are producing quantitative
information and some procedures that combine empirically- and physically-based methods have
been advanced. Vanoni (1975) presented six methods used for prediction of gully growth and/or
gully head advance. They all follow some type of multiplicative or power law and are replete
with empirical constants that are generally site specific. McCuen (1998) updated and further
described gully erosion prediction equations with the observation that five factors underlie the
relevant variables of the process: land use, watershed size, gully size, soil type, and runoff
momentum. Having investigated the relevant factors, however, McCuen found that none of the
equations treat all terms. Better methods of evaluating gully formation that are more physically
based are needed.
The potential for gully development in vegetated soil surface layers has been assessed at landfill
sites using the tractive force method described by Temple et al. (1987) and DOE (1989) and
developed for channel flow (see Section 2.2.5.5.2), the Horton/NRC method for computing the
critical distance for gully formation (NRC, 1990) (see Section 2.2.5.5.3), and the permissible
velocity method described by Chow (1959) and NRC (1990) and also developed for channel
flow (see Section 2.2.5.5.4). These methods are presented below and are based on the approach
of NRC (1990) guidance. This approach is to prevent gully initiation during the occurrence of a
single, extremely large, design rainfall. By designing for such an event, it is expected that
smaller, continual events will have little or no cumulative influence on gully initiation. Of
course, such a conservative approach results in relatively flat, and relatively short, slopes.
Similar approaches, typically using the permissible tractive force and velocity methods, can be
used to design other types of surface layers. For example, design methodologies for riprap
covering uranium mill tailings piles have been developed and used with apparent success.
Nelson et al. (1986) discuss general design methodologies, and Abt et al. (1988) present design
criteria based on flume tests. The NRC (NRC, 1990) recommends specific methodologies and
equations for the calculations. For example, the Stephenson method, described by Abt et al.
(1988) (see Section 2.2.5.5.5), can be used to select the mean particle diameter to withstand a
design storm. The Stephenson method is recommended for evaluating the erosion resistance of a
gravel or riprap layer with a slope inclination greater than 10% (NRC, 1990). For steeper slopes
(e.g., slope inclinations greater than 5H:1V), the Hartung and Scheuerlein method (Hartung and
Scheuerlein, 1970) has been used.
DRAFT - DO NOT CITE OR QUOTE 2-25
-------
2.2.5.5.2 Tractive Force Method for Vegetated Surface Layers
The tractive force method (Temple et al., 1987; DOE, 1989) can be used to calculate the
allowable shear stress, xa (kPa), of a vegetated surface layer as:
Ta = TabCe2>0.9kPa (Eq.2.6)
where: xab = allowable shear stress for the surface layer with bare soil (kPa); and Ce = void ratio
correction factor (dimensionless). Temple et al. (1987) and DOE (1989) provide graphs for both
Tab and Ce values.
The allowable shear stress must be equal to or greater than the effective shear stress applied to
the surface layer by the flowing water, xe (kPa):
(Eq. 2.7)
V n J
where: yw = unit weight of water (kN/m3); D = flow depth (m); S = slope inclination
(dimensionless); CF = vegetal cover factor (dimensionless); n = Manning's roughness coefficient
for the considered vegetative cover (dimensionless); and ns = Manning's roughness coefficient
for the bare soil (dimensionless). Guidance on the selection of values for the vegetal cover factor
and the Manning's coefficients is provided by Temple et al. (1987) and DOE (1989).
The depth of flow can be calculated using the Manning's equation (DOE, 1989):
D = fe] (Eq-2-8)
where: q = peak rate of runoff (m3/s/m) from Eq. 2.1 (and incorporating the flow concentration
factor), and all other terms are as defined previously.
2.2.5.5.3 Horton/NRC Method for Vegetated Surface Layers
The Horton/NRC method (NRC, 1990) is also used for prediction of gully formation for
vegetated surface layers. The method is used to estimate the critical distance, xc (m), along a
slope before gully formation begins. The slope lengths of a cover system should be designed to
be less than xc between runoff collection points (e.g., between drainage swales) to minimize the
potential for gully development. The equation for xc is as follows:
5/3
xc = ^- — (Eq. 2.9)
45Firn(f(S))5/3
where: Tah= allowable shear stress for the Horton/NRC method (kPa); F = flow concentration
factor (dimensionless) from Eq. 2.1; ir= rainfall intensity (m/s) from Eq. 2.2; n = Manning's
roughness coefficient for the considered vegetative cover (dimensionless), calculated using the
DRAFT - DO NOT CITE OR QUOTE 2-26
-------
tractive force method described in Section 2.2.5.5.1; and f(S) = slope function (dimensionless).
The allowable shear stress can be calculated as the minimum of:
(Eq.2.10)
Tah = Tva = 0.75C:
where: xa, CF, ns, and n are calculated using the tractive force method described in Section
2.2.5.5.1; Tva= limiting vegetal stress (stress at which vegetation will break) (kPa); and Q =
vegetal retardance curve index (dimensionless). Guidance on the selection of values for the
vegetal retardance curve index is provided by Temple et al. (1987) and DOE (1989). Eq. 2. 10 is
based on allowable soil stress, and Eq. 2.11 is based on allowable vegetal stress.
The slope function can be calculated as follows (NRC, 1990):
where: P = slope angle (degrees).
2.2.5.5.4 Permissible Velocity Method for Vegetated Surface Layers
The permissible velocity method (Chow; 1959; NRC, 1990) can also be used to assess the
potential for gullies to form in a vegetated cover. The flow velocity of runoff should be less than
the permissible velocity for the surface layer material. NRC (1990) recommends checking
results of the Horton/NRC Method against those of the permissible velocity method.
The flow velocity, v (m/s), is calculated in the conventional manner:
v = q/D (Eq. 2.13)
where all other terms are as defined previously.
Permissible velocities recommended by SCS (1986b) for a range of vegetated cover conditions
(e.g., grass type, surface layer slope, soil erosion sensitivity, etc.) in drainage channels are
presented in Table 2-3. When the flow depth, D, is less than 1 m, NRC (1990) recommends that
the permissible velocity in the channel be reduced by a reduction factor, Rf (dimensionless):
Rf = 1 + 0.46 log(D) for 0.08m
-------
Table 2-3. Permissible velocities recommended by SCS for vegetated drainage channels
(modified from SCS, 1986b).
Vegetation Type
Bermudagrass
Bahiagrass
Buffalograss
Kentucky bluegrass
Smooth brome
Blue grama
Tall fescue
Grass mixtures
Reed canarygrass
Lespedeza sericea
Weeping lovegrass
Yellow bluestem
Redtop
Alfalfa
Red fescue
Common lespedeza4
Sudangrass4
Slope Range
(%)
0-5
5-10
over 10
0-5
5-10
over 10
0-5
5-102
0-5J
0-5D
Permissible Velocity1
Erosion
resistant soils
(ft/s)
8
7
6
7
6
5
5
4
3.5
3.5
Easily
eroded soils
(ft/s)
6
5
4
5
4
3
4
3
2.5
2.5
1 Use velocities exceeding 5 ft/s only where good vegetated covers and proper maintenance can be obtained.
2 Do not use on channel slopes steeper than 10%, except for vegetated sideslopes in combination with a stone,
concrete, or highly resistant vegetative center section.
3 Do not use on channel slopes steeper than 5%, except for vegetated sideslopes in combination with a stone,
concrete, or highly resistant vegetative center section.
4 Use annuals on mild slopes or as temporary protection until permanent vegetated covers are established.
5 Use on slopes steeper than 5% is not recommended.
2.2.5.5.5 Stephenson Method for Gravel or Riprap Surface Layers
The Stephenson method (NRC, 1990) is used to compute the minimum gravel or riprap mean
particle diameter, DSQ (mm), to withstand the peak rate of runoff:
50
=1,000
q(tanp)//6n
1/6
Cd g172 [(l-njGs -l))cosp(tan^-tanp)J5/3
(Eq. 2.15)
where: np = porosity of gravel or riprap layer (dimensionless); Cd = empirical factor
(dimensionless) ranging from 0.22 for gravel to 0.27 for crushed granite (Stephenson, 1979); g =
acceleration of gravity (9.81 m/s2); Gs = specific gravity of gravel or riprap (dimensionless); § =
angle of repose of gravel or riprap (degrees); and all other terms are as defined previously.
Guidance on the selection of values for the porosity and angle of repose of the gravel or riprap is
provided by Abt et al. (1987) and NRC (1990). Gravel or riprap with a mean particle diameter of
DRAFT - DO NOT CITE OR QUOTE
2-28
-------
DSO will be on the threshold of movement under flow q. The surface layer will collapse at a flow
varying from 1.2q (for gravel) to 1.8q (for crushed granite) (Stephenson, 1979).
2.2.5.6 Wind Erosion
2.2.5.6.1 Revised Wind Erosion Equation.
The average annual rate of soil loss by wind erosion (for that portion of sediment that moves
between the soil surface up to a height of 2 m) can be estimated using the Revised Wind Erosion
Equation (RWEQ) computer program (Fryrear et al., 1998). RWEQ was developed for
agricultural fields and is currently being used by the NRCS to assess soil loss. The model is
derived from the theory of erosion processes and data from laboratory and field wind tunnel
studies.
Using finite difference techniques, RWEQ solves an equation for horizontal mass transport
across an eroding surface:
(Eq.2.16)
dx s(x)
where: Q(x) = mass transport of soil (kg/m) at downwind distance x; x = downwind distance
(m); Qmax(x) = maximum mass transport of soil (kg/m) at downwind distance x; and s(x) = field
length scale (m).
The maximum mass transport of soil, Qmax (kg/m), is calculated as:
Qmax =109.8 (WF EF SCF K' COG) (Eq. 2.17)
where: WF = weather factor (kg/m) and is a function of wind speed, soil wetness, snow cover,
and other factors; EF = erodible fraction (dimensionless), is the fraction of the surface 25 mm of
soil that is smaller than 0.84 mm, and is computed empirically as a function of the percentages of
clay, silt, and sand-sized particles, organic matter, and calcium carbonate in the soil; SCF = soil
crust factor (dimensionless) and is computed empirically as a function of the percentages of clay
and organic matter in the soil; K' = soil roughness factor (dimensionless) and is a function of soil
clod roughness, ridge height and spacing, and other factors; and COG = combined crop factors
(dimensionless) and is related to plant canopy and residues.
RWEQ uses monthly weather data, soils and field data, and management inputs to assess wind
erosion. The management inputs include cropping systems tillage and operation dates,
windbarrier descriptions, and irrigation information. Time periods from the management input
file are used to partition the weather factor for each management time period. The dominant
wind direction is assessed, and the wind factor is computed for four directions using weather data
and considering hill and wind barrier effects, snow cover, and soil moisture content. Operation
dates are also used to determine time periods for computation of residue decay, soil roughness
decline, and soil erosion. Residue decomposition is computed for each period based on weather
conditions and accumulated decomposition days since crop harvest. Soil roughness is decayed
DRAFT - DO NOT CITE OR QUOTE 2-29
-------
for each time period based on rainfall characteristics and clay content. The residue and soil
roughness for each time period are used with the length of eroding field to determine the average
soil erosion for that field length. The soil erosion from the different time periods are then
summed to get the average annual rate of soil loss by erosion.
Input values for RWEQ are developed using site-specific information and the database that is
part of the RWEQ computer program. The program is available for download from
http://www.csrl.ars.usda.gov/wewc/rweq/readme.htm.
2.2.5.6.2 Wind Erosion Prediction System
The Wind Erosion Prediction System (WEPS) is a process-based, daily time-step, computer
model that simulates weather, field conditions, and erosion. WEPS development involves an
Agricultural Research Service (ARS) led, national multidisciplinary team of scientists, intended
to replace the predominately empirical Wind Erosion Equation (WEQ) (Woodruff and
Siddoway, 1965). Agencies involved include the ARS, Natural Resource Conservation Service
(NRCS), and Forest Service (FS) from the U.S. Department of Agriculture, along with the EPA
and Bureau of Land Management (BLM). The purposes of WEPS are to improve technology
for assessing soil loss by wind from agricultural fields and to provide new capabilities such as
assessing soil movement, plant damage, calculating suspension loss, and estimating PM-10
(particles less than 10 microns in diameter) when wind speeds exceed the erosion threshold
(Wagner, 1996)
WEPS consists of an instructional program, a user-interface program, seven submodels, and an
output section. WEPS allows users to input their own data files or use previously prepared data
base files. It also possesses the ability to provide users with individual values for suspension,
saltation, and surface creep. WEPS' seven submodels, each based on the fundamental processes
which occur in the field, are used to predict and give estimates for wind erosion.
More information on WEPS and wind erosion can be found at the USDA-ARS Wind Erosion
Research Unit (WERU), available at http://www.weru.ksu.edu/.
2.2.5.7 Erosion Control Materials
One often-effective means for controlling erosion is through the use of erosion control materials.
Such materials can be temporary or permanent and, depending on the materials, are placed
before, during, or after seeding. Once installed, the measures may require maintenance to
maintain their effectiveness.
2.2.5.7.1 Temporary Erosion Control Materials
Temporary erosion and revegetation materials (TERMs) consist of materials that are in whole or
part degradable. TERMs provide temporary erosion control and are either disposable after a
given period, or only function long enough to facilitate vegetative growth. After the growth is
established, the TERMs are no longer needed. Some of the TERMs are completely
biodegradable, but others are only partially so. Theisen (1992) groups the various materials
listed in the upper part of Table 2-4 as being in the TERM category.
DRAFT - DO NOT CITE OR QUOTE 2-30
-------
The first two products listed in the TERM category in Table 2-4 consist of traditional methods of
erosion control using straw, hay, or mulch loosely bonded by asphalt or adhesive. The stability
of this type of material is may not be very good. Geofibers in the form of short pieces of fibers
or microgrids can be mixed into soil with machines or rototillers to aid in laydown and
continuity. The fiber or grid inclusions provide for greater stability over straw, hay, or mulch
broadcast over the ground surface.
Table 2-4. Erosion control materials (after Theisen, 1992)
Type of Material
Examples of Material
Temporary Erosion and Revegetation Materials
(TERMs)
Straw, hay, and hydraulic mulches
Tackifiers and soil stabilizers
Hydraulic mulch geofibers
Erosion control meshes and nets
Erosion control blankets
Fiber roving systems
Permanent Erosion and Revegetation Materials
(PERMs) - Biotechnical Related
UV-stabilized fiber roving systems
Erosion control revegetation systems
Turf reinforcement mats
Discrete length geofibers
Vegetated geocellular containment systems
Permanent Erosion and Revegetation Materials
(PERMs) - Hard Armor Related
Geocellular containment systems
Fabric formed revetments
Vegetated concrete block systems
Concrete block systems
Stone riprap
Gabions
Erosion control meshes and nets are biaxially oriented materials manufactured from
polypropylene or polyethylene. These materials do not absorb moisture, nor do they shrink or
expand over time. They are lightweight and are stapled to the seeded ground using hooked nails
or U-shaped pins. The purpose of affixing the material to the ground is to improve stability.
Erosion control blankets are also biaxially oriented nets or meshes manufactured from
polypropylene or polyethylene. With these materials, a blanket of straw, excelsior, cotton,
coconut, or polymer fiber is attached to one or both sides of the net or mesh. The fibers are held
to the net or mesh by glue, lock stitching, or other methods.
Fiber roving systems are continuous strands, or yarns, usually of polypropylene, that are fed
continuously over the surface to be protected. They can be placed by hand or using compressed
air. After placement on the ground surface, emulsified asphalt or other soil stabilizer is used for
controlled positioning.
2.2.5.7.2 Permanent Erosion Control Materials
Permanent erosion control materials (PERMs) can be biotechnical or hard armor (Table 2-4).
The biotechnical materials are discussed first.
DRAFT - DO NOT CITE OR QUOTE
2-31
-------
Most of the biotechnical materials are polymer products that control erosion, aid in vegetative
growth, and eventually become entangled with the vegetation to provide reinforcement to the
root system. As long as the material is shielded from sunlight, via shading and soil cover, it will
not degrade (at least within the limits of polymeric materials). The polymers can be stabilized
with carbon black and/or chemical stabilizers. The seed is usually applied after the PERM is
placed.
Erosion control revegetation mats and turf reinforcement mats are closely related materials, the
basic difference being that erosion control revegetation mats are placed on the ground surface
with a soil infill, while turf reinforcement mats are placed on the ground surface with soil filling
in and above the material. Thus, turf reinforcement mats can be expected to provide better
vegetative entanglement and longer performance. Seeding is usually done prior to installation of
an erosion control revegetation mat, but while backfilling within the structure of turf
reinforcement mats.
Discrete length geofibers are short pieces of polymer yarns mixed with soil to provide a tensile
strength component that can resist forces such as those occurring at athletic fields and on slopes.
Vegetated geocellular containment systems consist of three-dimensional cells of GMs or GTs,
which are filled with soil and vegetated (Figure 6-33).
Hard armor systems provide their own erosion protection, independent of vegetation.
Geocellular containment systems are permanent when the infill material is concrete. Fabric
formed revetments are GTs that are filled with concrete or grout. As the GT deteriorates over
time from UV degradation, the concrete or grout is left behind.
Numerous concrete block systems are available for erosion control. Hand placed interlocking
masonry blocks are popular for low traffic pavement areas such as driveways. The voids in the
blocks and between them are usually vegetated. Alternatively, the system can be factory
fabricated as a unit, brought to the job site, and placed on prepared soil. The prefabricated
blocks are either laid on, or bonded to, a GT substrate. The finished mat can bend and torque by
virtue of the blocks being articulated with joints, weaving patterns, or cables. A concrete
cribwall has also been used as a surface layer (Figures 6-30 and 6-31).
Stone riprap can be very effective as was discussed earlier. A GT placed on the soil surface
before placement of riprap serves as a filter and separator.
Gabions consist of discrete cells of wire netting filled with hand-placed stone. The wire is
usually galvanized steel hexagonal wire mesh, but in some cases can be a plastic geogrid.
2.2.6 Construction
If topsoil is used to construct the surface layer, the soil is only compacted nominally, if at all, to
facilitate plant root development. Even moderate amounts of compaction can result in decreased
root depth and density. As described by the NRCS (1996), compaction restricts rooting depth,
which reduces the uptake of water and nutrients by plants. It also decreases infiltration, which
increases runoff and, thus, erosion potential. To promote the growth of vegetation, it is generally
recommended that cover soils be placed at bulk densities less than the values given in Table 2-5.
DRAFT - DO NOT CITE OR QUOTE 2-32
-------
A gravel-soil mixture will require some compaction, but heavy compaction is neither necessary
nor desired. Rock riprap is normally placed loosely with little or no compaction. Where
asphaltic concrete has been used as the surface layer, road-paving equipment was used for
construction.
2.2.7 Maintenance
Maintenance is discussed in Chapter 9. The most important maintenance activities for the
surface layer involve maintaining the intended vegetative cover and the erosion control
measures, repairing erosion gullies, filling surface depressions caused by localized settlement,
and, as an associated activity, maintaining and repairing surface-water management structures.
Table 2-5. Minimum soil bulk density at which a root restricting condition may occur
(NRCS, 1996).
Soil Texture Bulk Density
(g/cm3)
Coarse, medium, and fine sand and
loamy sands other than loamy very fine sand 1.80
Very fine sand, loamy very fine sand 1.77
Sandy loam 1.75
Loam, sandy clay loam 1.70
Clay loam 1.65
Sandy clay 1.60
Silt, silt loam 1.55
Silty clay loam 1.50
Silty clay 1.45
Clay 1.40
2.2.8 Monitoring
Monitoring is discussed in Chapter 8. The surface layer should be monitored to identify
problems with excessive erosion, excessive differential settlement, or slope instability, assess the
health of the vegetative cover, and evaluate gas emissions, if gases are a concern. If the cover
system water balance is being assessed, the surface layer moisture content or matric potential and
surface-water runoff may also be monitored.
2.3 Protection Layer
The protection layer lies directly beneath the surface layer and, in some cases, may be combined
with the surface layer to form the "cover soil". The primary functions of the protection layer are
to protect the underlying cover system components and to temporarily store water that has
DRAFT - DO NOT CITE OR QUOTE 2-3
-------
percolated through the surface layer until it can be returned to the atmosphere by ET. The
underlying layers may need protection from erosion, exposure to wet-dry cycles, exposure to
freeze-thaw cycles, exposure to ultraviolet light, and biointrusion by plant roots, burrowing
animals, and humans. The storage of water in the protection layer provides a water reservoir to
support plant growth and reduces infiltration into underlying cover system components. The
protection layer may also serve to attenuate emissions of radon gas for those wastes that emit
radon.
2.3.1 General Issues
Occasionally, cover systems are designed without a protection layer. In such cases the surface
layer is placed directly on a drainage layer or hydraulic barrier. This design approach is usually
not recommended because erosion gullies may sometimes cut through the surface layer (if it is
relatively thin) and expose or even erode the underlying layers. The underlying layers may then
become damaged under prolonged exposure to the environment. For example, exposed CCLs
will usually develop desiccation cracks. As discussed in Section 7.2, even up to 0.75 m of cover
soil may not be sufficient to protect underlying CCLs from degradation. Geosynthetics are also
vulnerable to degradation from exposure to ultraviolet light. If the surface layer is vegetated
topsoil and there is no protection layer to provide stored water to plants, the vegetation may
experience excessive stress and even die when the topsoil moisture content decreases to low
levels. In most situations, the only justification for omitting the protection layer is if the
underlying layers require no protection and the surface layer is not vegetated.
With this in mind, the most important concerns with respect to the protection layer are generally
the level of protection required by the underlying layers and the water storage capacity required
to support any vegetation.
2.3.2 Elements of Design
Important questions that typically need to be addressed when considering the design of the
protection layer include:
• What materials are available to construct the protection layer?
• What thickness of protection layer material is needed?
• How should the protection layer be constructed?
• What type and frequency of maintenance should be employed?
• What type and frequency of monitoring should be employed?
2.3.2.1 Materials
The protection layer is usually constructed from on-site or locally available soil. As discussed in
Section 2.2.2.2.1, medium-textured soils, such as loams, have the best overall characteristics for
seed germination and the development of plant root systems. Fine-grained soils, such as silts and
clays, have excellent water-holding capability, which provides roots with water for plant growth
but limits the transport of oxygen to plant roots. In addition, fine-textured soils are vulnerable to
cracking when desiccated. Conversely, coarse-grained soils, such as sands and gravels, have low
water retention capacity and high saturated hydraulic conductivity. Coarse-grained soils can
DRAFT - DO NOT CITE OR QUOTE 2-34
-------
drain and dry out quickly, resulting in an insufficient moisture supply for plants. For example,
there have been instances in which cover soils at landfills became so dry that cover system
irrigation was required to maintain adequate soil moisture to support grass (post-closure
maintenance of a vegetative cover was required in the permits for the facilities). The addition of
water to the surface of a cover system is generally not recommended because one of the primary
purposes of a cover system is usually to limit infiltration of water into the underlying waste.
If a soil protection layer is placed above a drainage layer, filter criteria for the two layers should
be met. Filter criteria can be met in one of two ways: (1) ensuring that the materials themselves
meet the criteria (thus eliminating the need for a filter); or (2) installing a soil or GT filter at the
interface between the layers. Filters are discussed in Section 4.7.
If the primary role of the protection layer is to prevent biointrusion, cobbles, asphaltic concrete,
recycled concrete pavement, or similar materials are typically required. If both vegetative
support and preventing biointrusion are critical, the protection layer may consist of two or more
components, for example a layer of cobbles overlain by a GT filter and then a silty loam soil
layer.
2.3.2.2 Thickness
The required thickness of the protection layer depends on many factors including:
• need to protect underlying layers from damage due to wet-dry and freeze-thaw cycles;
• maximum depth of frost penetration;
• need to prevent accidental human intrusion, penetration by burrowing animals, or root
penetration into underlying materials;
• need to support vegetative growth by accommodating plant roots;
• need to temporarily store water in the protection layer to attenuate rainfall infiltration into
the underlying layers and to sustain vegetation through dry periods;
• need to provide other types of protection unique to a particular waste (e.g., attenuate
radon emissions if the underlying waste emits radon); and
• need for a capillary barrier (discussed in Section 3.3), if this is a design strategy.
As previously mentioned in Section 2.2.2.3, thicknesses of cover soils (surface layer plus
protection layer) are often in the range of 0.45 to 0.6 m, although thicknesses greater than 1 m
are sometimes necessary to provide adequate rooting depth, soil moisture storage capacity, and
freeze-thaw protection or to meet other design requirements. The protection layer may need to
be still thicker if both vegetative support and protection from biotrusion is required. As will be
subsequently discussed, the typical thickness of a biointrusion-resistant cobble layer is on the
order of 0.5 to 1 m.
2.3.2.2.1 Desiccation Protection
Depending on the cover system components, the protection layer may need to be designed to be
thick enough to protect the underlying layers from desiccating. For example, the hydraulic
DRAFT - DO NOT CITE OR QUOTE 2-35
-------
integrity of a CCL will be compromised if it is allowed to desiccate and crack after being
exposed to wet-dry and/or freeze-thaw cycles. The degree of desiccation protection required for
a CCL depends upon whether the CCL is covered with a GM. If the barrier is a GM/CCL
composite, the GM will provide the CCL with some protection from desiccation (see Section
7.2). However, a soil protection layer with a thickness on the order of 0.45 m or more is still
required over the GM.
If the hydraulic barrier is a CCL alone, the problem of protecting the CCL from desiccation is
particularly challenging. As discussed in Section 2.5.2.6, cover soils have exhibited severe
desiccation to depths of up to 1 m, and possible deeper. It thus appears that the thickness of
protection layer required to slow desiccation of an underlying CCL that is not covered with a
GM for a time period of 30 years or more is at least 1 m, and probably more. Because only
limited information is available on this subject, a conservative approach is recommended.
Depending on the chemistry of the permeating water, GCLs may or may not be vulnerable to
permanent damage from desiccation (see Section 2.5.2.6). If the permeant contains cations that
may exchange with the sodium in the GCL bentonite, the barrier will loose some capability to
swell and recover from desiccation over time. As described in Section 2.5.2.6, GCLs have been
damaged for this reason in at least several field installations.
If it is desired to protect a CCL, GCL, or other type of barrier from desiccation (and it almost
always is desired to do so), the best approach is to place a GM over the barrier, and then cover
the GM with soil.
2.3.2.2.2 Frost Penetration Protection
The protection layer is generally designed with the intent of preventing underlying layers from
freezing. This is especially a concern in northern climates. As temperatures drop and soil layers
within the cover system freeze, water drawn towards the freezing front can cause desiccation
cracking, freeze-thaw cracking, and frost heaving. As discussed in Section 2.5.2.7, desiccation
and frost cracking may cause CCLs located within the frost zone to have increased permeability
to water and gas. Neither GCLs nor GMs appear to be vulnerable to freeze-thaw damage.
However, based on the information presented in Section 2.3.2.2.1, if freezing temperatures cause
a GCL to desiccate, it may become damaged if it rehydrates with water containing certain
exchangeable cations. To avoid damage to a CCL, the protection layer and overlying surface
layer should be thick enough to place any CCL below the maximum depth of frost penetration.
If may be advisable to also use this approach for GCLs. Alternatively the GCL may be covered
with a GM to reduce its potential to desiccate due to freezing conditions.
The protection layer should generally prevent the drainage layer (if one is present) from freezing
as well, particularly on relatively steep sideslopes. If the drainage layer freezes, it is not
functional for part of the year. During the thaw period, it is particularly important that the
drainage layer work properly, i.e., drain freely, and that the protection layer be sufficiently thick
to provide the protection that is required. If the drainage layer is to be within the depth of frost
penetration, the layer should be made permeable enough that it drains rapidly and has little
capillarity (i.e., has a low field capacity) so that the voids in the layer are filled with air and not
water during the winter months.
DRAFT - DO NOT CITE OR QUOTE 2-36
-------
The depth of frost penetration in a cover system may vary from that of the native deposits due to
differences in soil texture, moisture content, density, organic matter, and other factors. For
example, because clay particles have a higher insulation value than silt or sand particles and
since clay soils normally hold more moisture than silts and sands, the depth of frost penetration
is usually greater in silt and sandy soils (light-textured soils) than in clays and silty clays (heavy-
textured soils).
There are several techniques available for estimating the depth of frost penetration. One
common practice is to use frost penetration maps for native soils, such as the one in Figure 2-9.
This map shows contours of maximum frost penetration depth based on estimates made by the
U.S. Weather Bureau. Frost penetration maps may be of limited accuracy. According to
DeGaetano et al. (1997), available maps for maximum frost penetration depths in the U.S. are
based on unofficial, poorly documented, and antiquated (1899-1938) measurements.
N
Figure 2-9. Contours of Maximum Frost Penetration Depth (mm) and State Averages
(mm) (modified from Koerner and Daniel, 1997).
As an alternative to using frost penetration maps, the depth of frost penetration may be computed
using the air freezing index and other site-specific factors. The air freezing index is the total
number of degree-days of freezing for a given winter. One degree-day of freezing results when
the mean air temperature measured at 137.3 cm above the ground for one day is IF degree below
32°F. Air freezing index data and statistics (based on 1951-1980 data) for a number of weather
stations across the U.S. can be downloaded from the National Climatic Data Center (NCDC)
DRAFT - DO NOT CITE OR QUOTE
2-37
-------
website (http://lwf.ncdc.noaa.gov/oa/fpsf/fpsf.html); data documentation for the air freezing
index statistics is presented by Steurer (1998). The NCDC website also includes a map of 100-
year return period air freezing indices (Figure 2-10). There are a number of semi-empirical and
physical models for evaluating the frost penetration depth using the air freezing index. The most
commonly used model to evaluate the frost depth is the modified Berggren method. This semi-
empirical method, which is not presented in this guidance document, considers the thermal
properties of the soil layers, the air freezing index, and other parameters. Information on the
Berggren method can be found in Aldrich and Paynter (1953).
AIR-FREEZING INDEX (°F-DAYS)
ESTIMATE OF THE 100-YEAR RETURN PERIOD
Figure 2-10. Contours of Air Freezing Indicies (°F-days) with a 100-yr Return Period
(downloaded from http://lwf.ncdc.noaa.gov/oa/fpsf/fpsf.html).
2.3.2.2.3 Accidental Human Intrusion Protection
Accidental human intrusion has generally not been a design consideration for cover systems on
most landfills or waste remediation sites. However, ordinary human activities can damage the
cover system. For example, ruts may be created if vehicles are driven on the cover system when
the surface layer is wet. Normally, if an adequate cover soil thickness is provided to support
vegetation and protect the underlying cover system components, the thickness will also be
sufficient to protect the cover system from ordinary human impacts such as vehicle ruts.
DRAFT - DO NOT CITE OR QUOTE
2-38
-------
Essentially the only type of waste for which accidental human intrusion has been a design
consideration is radioactive waste. It is not clear why radioactive waste has been singled out.
Human intrusion into MSW or HW could also be dangerous to the intruder. When human
intrusion has been considered, the principal concern has been with accidental exposure (e.g.,
excavation to lay a buried pipeline or to construct a basement for a home). Though the cover
system can be thickened to approximately 5 m or more to prevent such occurrences, the problem
is more typically handled by assuming that deed restrictions and security measures will prevent
intrusion. No amount of thickness can prevent "intentional" intrusion, such as drilling a boring
or digging a deep utility excavation.
Some cover systems, especially those at redeveloped sites, may incorporate visible barriers with
bright, readily identifiable colors within or beneath the protection layer to indicate that the cover
system may be damaged if the intrusive activity continues any further downward. For example,
bright orange plastic netting has been used for such a purpose. Other types of visible barriers
may also be used to provide an additional safeguard against accidental digging or other
construction-related damage to the cover system.
2.3.2.2.4 Root Penetration Protection
The penetration of plant roots below the protection layer is undesirable. Suter et al. (1993)
summarize the potential mechanisms by which plant roots can damage a cover system:
• Roots may enter the drainage layer or gas collection layer and cause clogging.
• Roots may penetrate the hydraulic barrier, causing an increase in hydraulic conductivity.
• Decomposing roots leave channels for movement of water and vapors.
• Roots may desiccate CCLs, causing shrinking and cracking.
• Uprooted trees may lead to soil erosion and leave depressions in the cover system.
• Roots may enter the wastes, take up constituent chemicals, and transport them to above
ground components. For radioactive wastes, this is a particular concern.
• Roots may modify the waste by increasing decomposition rates and by releasing
chemicals that mobilize metals.
Suter et al. (1993) provide examples of several of these potential problems. Different plant
species develop root systems that penetrate to different depths. Root systems of shallow-rooted
grasses may penetrate no deeper than 0.15 m into the subsoil. Grasses with deeper root systems
may have roots that penetrate to depths of 0.3 to 0.5 m. Root systems of shrubs can penetrate to
depths in excess of 1 m. Some desert plant species have roots that can penetrate many meters
into the subsurface. Trees also have deeper root systems. In generally, the establishment of
deep-rooted shrubs and trees on a cover system should be prevented via routine maintenance
such as periodic mowing unless the cover system has been specifically designed to accommodate
the deep roots.
Climate influences the depth of root penetration, and even the materials into which roots
penetrate have an influence on root depth. Roots generally seek out lightly-compacted soils that
contain moisture. Roots will not, as a general rule, penetrate into dry or heavily compacted soils.
DRAFT - DO NOT CITE OR QUOTE 2-39
-------
In soil profiles containing a finer-grained soil overlying a coarser-grained soil, roots will remain
in the relatively moist, finer-gained soil and will not penetrate into the coarser-grained soil as
long as the coarser soil remains dry. If the coarser-grained soil becomes wet, then the roots will
seek moisture in this soil.
The coarser-grained material used to construct a barrier to plant roots often consists of cobbles.
When cobbles are used as a barrier to plants roots, the placement of a fine-textured soil over the
cobbles will create a capillary barrier. If the cobbles remain dry, they should stop further
downward penetration of plant roots (Hakonson, 1986). The cobbles may also help increase
plant growth by keeping moisture on the upper soil layer. Experiments with cobble biobarriers
have been carried out at arid and semi-arid sites (Cline, 1979; and Cline et al., 1982). Research
indicates that 0.9 m of cobbles, or 0.15 m of gravel over 0.75 m of cobbles, is effective in
stopping root penetration of deep-rooted plants (DePoorter, 1982).
Another alternative is to utilize materials that inhibit root growth, to stop further penetration of
roots into the soil. Cline et al. (1982) examined the effectiveness of several phytotoxins
impregnated into or onto GTs that were placed within the soil protection layer, just above the
drainage layer. Some of the phytotoxins met the goal of being effective in stopping the
downward progress of root growth, with no other effects. However, some of the phytotoxins
killed the plants when the roots encountered the fabric. The longevity of these products requires
further evaluation.
2.3.2.2.5 Burrowing Animal Protection
For some types of waste (particularly radioactive waste), the protection layer may need to
provide the cover system with a high level of protection from intrusion by burrowing animals.
Suter et al. (1993) summarize the effects that burrowing animals can have on cover systems as
follows:
• Animals may burrow through the cover system, resulting in direct channels for
movement of water, vapors, roots, and other animals.
• Even when they do not penetrate the entire cover system, burrows may increase the
porosity of the soil, thereby increasing infiltration rates in some situations (although, in
arid areas, burrows may actually do the opposite by provide channels for enhanced
evaporation).
• If burrows penetrate the entire cover system, animals may become externally
contaminated or consume the waste, thereby spreading the waste in their feces, urine, and
flesh.
• Animals may carry waste directly to the surface during excavation if the burrows fully
penetrate the cover system.
• By working the soil and transporting seeds, burrowing animals may hasten the
establishment of deep-rooted plants on the cover system.
• Burrowing animals cast soil on the surface, thereby increasing erosion of the cover
system.
DRAFT - DO NOT CITE OR QUOTE 2-40
-------
Research by Cline (1979), Cline et al. (1982), and Hakonson (1986) found that if objects, such as
cobbles, placed in a burrowing animal's path are sufficiently large and/or tightly packed, the
animal's progress is effectively stopped. Thus, a barrier to burrowing animals typically consists
of a 0.5 to 1-m thick layer of cobbles. The maximum particle size should be established based
on the burrowing animals of concern but is typically on the order of 100 to 200 mm. Care should
be taken to provide adequate filter layers both above and below the cobbles, to prevent overlying
and underlying soil particles from migrating into the cobbles. Filter design is presented in
Section 4.7.
A GM may also be viewed as a barrier to burrowing animals. Studies indicate that animals will
not make their way through GMs such as those made from HDPE (Steiniger, 1968). Also,
welded wire mesh and certain polymeric erosion control mats may also be barriers to burrowing
animals.
2.3.2.2.6 Vegetation Support
Vegetated cover soils should be thick enough to accommodate a healthy growth of plant roots
and store sufficient water to support plant growth. Plants should generally have relatively
shallow roots so that the roots do not penetrate too deep into the cover system because, as
described in Section 2.3.2.2.4, deep penetration threatens the integrity of underlying components.
However, roots should be deep enough to enable the plants to extract moisture from a sufficient
depth. Most grasses are thought to have effective rooting depths of about 0.15 to 0.5 m. If
plants with deeper roots are planted or represent a desirable climax community, the thickness of
the cover soil should be increased to accommodate root growth. For example, deeper-rooted
plants may become established over time and displace the grasses that were initially planted.
The minimum thickness of the cover soil is typically 0.45 to 0.6 m to accommodate plant roots.
Even thicker cover soils are required to accommodate certain shrubs and desert plant species.
2.3.2.2.7 Water Storage
Most of the rainfall that contacts the surface of a cover system infiltrates into the underlying
cover soil and is retained in the soil by capillary forces. The ultimate fate of this water is
primarily ET. For cover systems with a vegetated surface layer, it is critical that the cover soils
be capable of retaining sufficient moisture to support plant growth.
The greater the percentage of fines in a soil, the greater the water retention after gravity drainage.
The volumetric water content of a soil after gravity drainage is referred to as the soil's field
capacity, 0fc (dimensionless). This parameter is often reported as the volumetric water content at
a matric potential of -0.03 MPa (-3.3 m). At water contents less than field capacity, the soil
hydraulic conductivity is often assumed to be so low that gravity drainage of the soil becomes
negligible and the soil moisture is held in place by capillarity. Some of this stored water can be
removed via transpiration. Vegetation can reduce the soil moisture content from field capacity to
wilting point, 9wp (dimensionless). This parameter is often defined as the volumetric water
content at a matric potential of-1.5 MPa (-150 m)). At water contents below the wilting point,
plant activity is assumed to stop. Evaporation from the soil surface can further reduce the soil
moisture content from wilting point to residual saturation, which is the water content at an
DRAFT - DO NOT CITE OR QUOTE 2-41
-------
infinite matric potential. The relationship between these different soil water contents is shown in
Figure 2-11 for soil textures ranging from sand to clay.
Though plastic clays have a high field capacity, they are typically not used for the protection
layer because they can desiccate and crack, providing preferential pathways for infiltrating water
to bypass the clay matrix and thereby bypass storage. In addition, there is less water storage for
plants in these soils than in silty loam soils, as shown in Figure 2-11 and Table 2-6. In some
regions, such as the Texas Gulf coast, the surface soils are almost entirely highly plastic clays.
In such cases, there may be no practical alternative to the use of a heavy clay soil. If a loamy
soil is available, it is usually selected because it is the best soil in terms of combining good
moisture retention, workability, resistance to desiccation cracking, and moderate hydraulic
conductivity. Sandy clays, clayey sands, and lean clays may also be suitable for use in
protection layers.
0.60
0.50
0.40
K
Z
LJJ
^ 0.30
O
O
[Jj 0.20
0.10
0.00
POROSITY
WATER
STORAGE
CAPACITY
IPHM SATURATION
SAND SANDY LOAM SILTY CLAY SILTY
LOAM LOAM LOAM CLAY
CLAY
Figure 2-11. Relation Among Moisture Retention Parameter and Soil Texture Class
(modified from Schroeder et al., 1994).
A soil's available water storage capacity (i.e., 9fc - 0wp) depends on its texture and density.
Representative moisture content values for soils of different textures are given in Table 2-6.
Since cover soils are only lightly compacted (unlike hydraulic barriers which are heavily
DRAFT - DO NOT CITE OR QUOTE
2-42
-------
compacted), only data for low-density soils are presented. As shown in the table, silty or clayey
sands, silts, and silty clays typically have a storage capacity of about 0.1 to 0.15.
The depth of water, Hw (m), that can be stored in a soil layer for subsequent removal by plants
can be calculated as follows:
Hw — 0SC Hs — (0fc - 0wp) Hs
(Eq. 2.18)
where: 9SC = water storage capacity of soil (dimensionless); Hs = soil layer thickness (m); and all
other terms are as defined previously. It is important to note that the use of field capacity and
wilting point is arbitrary and ignores other factors that affect the amount of moisture retained in a
soil layer, such as rock fragments and salts in solution (Cassel and Nielsen, 1986; NRCS,
1998b). Nevertheless, these are simple and commonly used concepts and are applicable for
approximating the water storage capacity of a soil layer.
Table 2-6. Representative water contents for low-density soils with different textures
(modified from Schroeder et al., 1994).
Soil
Description
Clean, poorly-
graded sand
Clean, well-
graded sand
Silty sand
Low-plasticity
silt
Low-plasticity
silt
Low-plasticity
clay
Clayey sand
High-plasticity
clay
US DA
Classification
Coarse sand
(CoS)
Fine sand
(FS)
Sandy loam
(SL)
Loam
(L)
Silty loam
(SiL)
Clay loam
(CL)
Sandy clay
(SC)
Clay (C)
Porosity
(-)
0.417
0.457
0.453
0.463
0.501
0.464
0.430
0.475
Field
Capacity
(-)
0.045
0.083
0.190
0.232
0.284
0.310
0.321
0.378
Wilting
Point
(-)
0.018
0.033
0.085
0.116
0.135
0.187
0.221
0.251
Storage
Capacity
(-)
0.027
0.050
0.105
0.116
0.149
0.123
0.100
0.127
Saturated
Hydraulic
Conductivity
(mis)
1.0x10"4
3.1 x10"D
7.2x10"b
3.7x10~D
1.9x10"D
6.4x10"'
3.3x10"'
2.5x10"'
The depth of water that can be stored in a soil layer can be substantial. For example, from Table
2-6 and Eq. 2.18, the representative storage capacity of a 0.6-m thick protection layer constructed
with silty loam is 0.149 and the depth of water that can be stored in this layer is approximately
90 mm. If the protection layer was constructed with fine sand, only about one-third of this
storage capacity would be provided.
DRAFT - DO NOT CITE OR QUOTE
2-43
-------
2.3.2.2.8 Radon Attenuation
Some radioactive wastes emit radon-222 (222Rn) in the form of a heavier-than-air gas. Inhalation
of radon gas at sufficient concentrations is a human health hazard. Federal regulations limiting
radon releases to the atmosphere are contained in 40 CFR §192.02 and are applicable to the
control of emissions from UMTRA sites that must comply with UMTRCA. The regulations are
also typically applied as an ARAR to DOE sites undergoing remediation. These regulations
require that release of 222Rn to the atmosphere not exceed: (i) an average release rate of 20
picocuries per square meter per second; or (ii) increase the annual average concentration of 222Rn
in the air at or above any location outside of the disposal site by more than one-half picocurie per
liter. To attenuate the release of radon to the environment, the cover system may need to
incorporate a radon gas barrier. This barrier may be incorporated in the hydraulic barrier or it
may be located closer to the surface, in which case the gas barrier may be considered to be part
of the protection layer.
GMs can also be used as barriers to radon gas release. While the half-life of 222Rn is short (3.8
days), radon is a part of the uranium-238 (238U) decay series. Uranium-238 has a half-life of
about 4.5 billion years. Given this long half-life, there has been some concern about the
longevity of GM barriers used for radon control. Although GMs will not last forever, a properly
selected and appropriately formulated GM, adequately protected by design, can last for a
presumed timeframe measured in hundreds of years. Because the cost of GMs is relatively low,
a GM can provide a cost-effective means of radon gas control for the timeframe just indicated.
For a soil layer to function as an effective barrier to gas diffusion, air-filled voids in the soil have
to be discontinuous. Gas diffuses very slowly through wet soils that contain only occasional,
unconnected air bubbles. Relatively thick (up to about several meters) layers of clay-rich soil are
typically employed when protection from radon emissions is needed. For clayey soils to
function effectively as gas barriers, they must be at a high degree of saturation and free of cracks.
Over a design life of hundreds of years, maintaining a wet, undesiccated layer of clayey soil
under natural conditions can be a tremendous challenge. To maintain a high water content in the
soil, a riprap surface layer may be considered to increase infiltration. The increased infiltration
may, however, result in increased potential for percolation through the cover system.
Specific procedures for designing soil layers to provide radon protection are beyond the scope of
this guidance document. One methodology documented by DOE (1989) involves determining
the allowable radon emission, estimating the radon diffusion coefficient through the soil, and
sizing the thickness of the soil layer based on the calculated diffusive flux. Additional
information on radon attenuation through cover systems is presented in NRC publications by
Rogers and Associates Engineering (1984a,b).
2.3.3 Construction
When the cover system is vegetated, the soil protection layer is only lightly compacted to allow
plant roots to penetrate the soil, as discussed in Section 2.2.6. For unvegetated cover systems,
the soil protection layer may be placed and compacted using procedures for structural fill or may
have no specific compaction criteria. Depending on the properties of the materials underlying
the protection layer, and especially if there are geosynthetics underlying the protection layer,
there may be limitations on the stresses exerted by the construction equipment. For example, if a
DRAFT - DO NOT CITE OR QUOTE 2-44
-------
soil protection layer overlies a GC drainage layer, the soil may need to be placed with a low-
ground pressure bulldozer and a minimum first lift compacted thickness of 0.2 to 0.3 m.
2.3.4 Maintenance
Maintenance is discussed in Chapter 9. Since the protection layer is covered by the surface
layer, protection layer maintenance is generally not needed unless the surface layer is breached
due to erosion or there are problems with excessive differential settlement or slope instability.
2.3.5 Monitoring
Monitoring is discussed in Chapter 8. If the cover system water balance is being assessed, the
protection layer moisture content or matric potential may be monitored.
2.4 Drainage Layer
Water that permeates through the surface and protection layers can be removed from the cover
system by an internal drainage layer. The primary functions of the drainage layer are to: limit
the buildup of hydraulic head on the underlying hydraulic barrier, which minimizes percolation
of water through the barrier; drain the overlying protection and surface layers, which increases
the available water-storage capacity of these layers and helps to minimize erosion of these layers;
and reduce the seepage forces in the protection, surface, and drainage layers, which improves
cover system slope stability.
2.4.1 General Issues
In many cases and especially on sideslopes, an internal drainage layer is included above the
hydraulic barrier to promote lateral drainage and prevent the buildup of hydraulic head in the
cover system. As discussed by Bonaparte et al. (2002), the design of existing cover system
drainage layers has been found to be inadequate in a significant number of cases, leading to a
significant number of instances of excessive cover system erosion and slope instability. The
main issues with drainage layer design are related to flow capacity, transitions and outlets, and
filtration. Each of these issues is discussed below.
The drainage layer should be designed to have adequate flow capacity. As described in Section
7.4.3, there have been cases of cover system instability due to the build up of seepage forces on
sideslopes after a rainfall. For some of these cases, the drainage layer was not designed with
adequate flow capacity; in one case, the cover system did not include a drainage layer. The
drainage layer should be designed to convey the maximum anticipated flow rate from a design
storm, and the maximum flow rate should be calculated considering the cover system water
balance for the selected storm. Methods for calculating the maximum flow rate in a drainage
layer are presented in Section 4.5. The allowable flow rate of a drainage layer can be calculated
as described in Section 2.4.2.3.
It is noted that in arid and semi-arid climates a water balance may show that a cover system does
not require a drainage layer. Instead, it may show that infiltration is stored in the overlying cover
soils and later removed by ET.
DRAFT - DO NOT CITE OR QUOTE 2-45
-------
Drainage layer transitions and outlets should be designed to provide free-flow of water.
Otherwise, cover soils can become saturated, leading to increased erosion, and seepage forces
can increase, leading to an increased potential for slope instability. The design of drainage layer
slope transitions is discussed in Section 4.6. Outlet design is discussed in Section 2.4.2.4.
The need for a soil or GT filter above the drainage layer should be evaluated. Sometimes the
drainage material (particularly if it is sand) is inherently a filter for the adjacent materials, in
which case a separate filter layer is not required. However, a filter (soil or GT) is usually
required, particularly if the drainage layer is gravel or a GN. As described in Section 7.4.3, there
have been cases of cover system instability where the cause of the instability was attributed to
clogging of a GT filter or clogging of a granular drainage layer when a filter layer was omitted.
If a filter is required, it should be designed to retain the overlying soil, resist clogging, and have
adequate permittivity. The design approach for soil and GT filters is presented in Section 4.7.
2.4.2 Elements of Design
Important questions that typically need to be addressed when considering the design of the
drainage layer include:
• What materials are available to construct the drainage layer?
• What thickness of drainage layer material is needed?
• What are the maximum design flow rate and allowable flow rate in the drainage layer?
• How should drainage layer transitions and outlets be designed?
• How should the drainage layer be constructed?
• What type and frequency of maintenance should be employed?
• What type and frequency of monitoring should be employed?
2.4.2.1 Materials
Both granular materials (typically sand or gravel) and geosynthetics (GT, GN, and GC) have
been used as drainage layer material in cover systems. The material used should have adequate
hydraulic conductivity to minimize the buildup of hydraulic head above the hydraulic barrier and
adequate hydraulic transmissivity to convey the design flow rate. The drainage layer material
should also meet filter criteria with adjacent layers.
2.4.2.1.1 Granular Materials
Granular drainage materials are normally composed of relatively clean sand or gravel. Gravel is
material that does not pass through the 4.74-mm wide openings of a No. 4 sieve. Sand consists
of material that passes through the No. 4 sieve but not through the 0.075-mm wide openings of a
No. 200 sieve. "Clean" sand or gravel refers to sand or gravel that contains very little or no
material that passes through the openings of a No. 200 sieve. Clean sands and gravels are often
produced by washing natural sands and gravels to remove any "fines," which are particles that
pass through the openings of a No. 200 sieve.
DRAFT - DO NOT CITE OR QUOTE 2-46
-------
The drainage layer should meet filter criteria with the overlying protection layer. If the drainage
layer material will not retain the protection layer material, a soil or GT filter is required. A
discussion of filter layer design is presented in Section 4.7.
Specifications for granular materials often require:
• no more than 5% (dry-weight basis) of material passing the No. 200 sieve;
• a maximum particle size on the order of 25 to 50 mm; however, smaller particles will
typically be required if a GM will underlie the drainage layer; alternatively, a GT cushion
layer can be used;
• restrictions on gradation, stated in terms of allowable percentages for specified sieve
sizes (these restrictions may exist for various purposes, including filtration
considerations);
• limitations on mineralogy (often the drainage material is required to be a non-
carbonaceous material, with a limit on the amount of calcium carbonate in the material,
although hard evidence that carbonaceous materials are truly unsuitable is lacking, as
discussed below);
• restrictions on the angularity of the material, if the material will interface with
geosynthetics, which are vulnerable to puncture by large, sharp objects (or, alternatively,
a GT cushion may be employed);
• that no deleterious material be present; and
• a minimum acceptable saturated hydraulic conductivity.
The specified material requirements attempt to ensure that the materials will not puncture
adjacent geosynthetics, will be chemically stable, and will provide adequate drainage. Perhaps
the two most complex requirements relate to presence of calcium carbonate and to hydraulic
conductivity.
Nearly all granular construction materials are natural, excavated materials (e.g., river sand or
gravel) or are produced from crushing rock. In either case, granular materials that are rich in
calcium carbonate (e.g., crushed limestone or dolostone) are commonly available in many parts
of the U.S. and are frequently considered for use as drainage layer material. There are two
concerns over the use of drainage material containing calcium carbonate. First, if GCLs are used
as the hydraulic barrier, teachable calcium may undergo ion exchange with the sodium in the
bentonite causing an increase in the GCL's hydraulic conductivity. (CCLs can also be adversely
impacted by ion exchange, but generally to a much lesser extent because of their thickness and
minerology.) Second, calcium carbonate may slowly dissolve, threatening the integrity of the
drainage material and potentially causing chemical clogging if the dissolved material is
precipitated elsewhere in the system. There is little hard, published evidence that dissolution of
calcium carbonate from drainage materials in cover systems is, in fact, a serious problem.
However, the mechanism is obvious and the potential for problems commands caution. This is
an area of on-going research, and, within the next few years, it should be possible to develop
additional design guidance. However, until more definitive information becomes available, it is
recommended that the calcium carbonate content of the drainage material be limited.
DRAFT - DO NOT CITE OR QUOTE 2-47
-------
Although there are no definitive guidelines, specified maximum values for calcium carbonate
content typically range from 5 to 20%. Local experience and practice, coupled with knowledge
of the calcium carbonate content of locally available granular materials, tend to dictate the
specified value. In some areas, it may be impossible to find granular materials that are
completely free of calcium carbonate. In addition, of the two ASTM tests that are often specified
for calcium carbonate content (ASTM D 3042 and ASTM 4373), one has been criticized for not
providing reproducible or reliable test results for granular drainage materials and both use strong
acids to dissolve the calcium carbonate.
No specific minimum hydraulic conductivity is recommended for a granular drainage material
because the required value is site dependent. When there is a regulatory guidance or requirement
(e.g., Federal guidance regarding cover system drainage layers for HW landfills), the minimum
specified hydraulic conductivity is generally 1 x 10"4 m/s. However, analysis indicates that this
value may be too low for many applications. The problem with a minimum hydraulic
conductivity of 1 x 10"4 m/s is that it may not provide the drainage layer with sufficient capacity
to convey the maximum flow rate from a design storm. To minimize the potential for excessive
erosion and slope instability, the drainage layer should be able to convey the maximum flow rate
entirely in the layer without buildup of excess head.
Also, a soil with a hydraulic conductivity of 1 x 10"4 m/s will typically retain a significant
amount of moisture under gravity drainage conditions (i.e., have a significant field capacity).
The presence of this moisture increases the potential for root penetration into the layer. The
moisture also increases the potential for freeze-thaw effects.
Hydraulic conductivity is usually measured in the laboratory using ASTM D 2434. The degree
of difficulty in accurately measuring hydraulic conductivity increases as the hydraulic
conductivity increases. With very high-hydraulic conductivity materials (e.g., large gravels), it is
necessary to maintain a very low head loss in order to avoid turbulent flow, and the small head
loss is difficult to measure. Specialized laboratory equipment is required to test these materials.
Care should be taken to ensure that representative samples of material are tested for hydraulic
conductivity, and that the density (hence, porosity) of the samples are representative of the value
expected for the drainage layer as constructed in the field. As materials are handled in the field,
they tend to get ground up slightly, producing additional fines and lowering hydraulic
conductivity, particularly in the lower part of the drainage layer. As a rule of thumb,
approximately 0.5 to 1% of additional fines by weight will be generated every time a drainage
material is handled. When a sample is collected from a material stockpile, there is a tendency to
select a sample near the surface. Such samples may be cleaner than material from deeper in the
stockpile and also cleaner than the material will be after it is handled and placed in the field.
2.4.2.1.2 Geosynthetics
Because the normal stresses on a cover system drainage layer are relatively low, a number of
different types of geosynthetics can be considered for use as the drainage layer. Geosynthetic
drainage materials most frequently used in cover systems include:
DRAFT - DO NOT CITE OR QUOTE 2-48
-------
• GNs of solid ribs with diamond-shaped apertures;
• GNs of foamed ribs with diamond-shaped apertures; and
• needlepunched nonwoven GTs.
Other geosynthetics drainage materials that may also meet project-specific requirements include:
• "high flow" GNs of solid ribs in a parallel orientation;
• drainage cores of single cuspations or dimples;
• drainage cores of double cuspations or dimples;
• drainage cores of built-up columns;
• drainage cores of stiff three-dimensional entangled mesh;
• resin bonded nonwoven GTs.
Like granular drainage layers, a geosynthetic drainage layer should meet filter criteria with the
overlying protection layer. A GN or core drainage layer requires an overlying GT filter to keep
the protection layer material from directly clogging the apertures of the drain. Furthermore, if a
GM hydraulic barrier underlies a GN or core drainage layer, as is often the case, a GT may be
required between the drain and GM to provide higher interface friction on steep sideslopes and,
possibly, reduce deformation-related intrusion of the GM into the drain and/or protect the GM
from puncture or other damage by the drain. Often, the GT is heat bonded or glued to the GN or
drainage core, creating a GC, to enhance interface shear strength, decrease the potential for
fugitive soil particles to enter the drain during construction, and facilitate installation. If a GT
drainage layer is used, it is also designed to meet filter criteria with the overlying protection layer
material.
A potential advantage of thin geosynthetic materials as drainage layers is that the weight of these
materials is very low, which is advantageous when compressible waste or soil underlies the
cover system. Also, geosynthetics, being thin, occupy less airspace than an equally transmissive
granular drainage layer. (This same advantage applies to the use of a GCL over a CCL as a
hydraulic barrier and a geosynthetic over granular material in a drainage layer.)
Specifications for geosynthetic drainage layers often require:
• resin and additive requirements;
• minimum thickness;
• minimum mass per unit area;
• minimum hydraulic transmissivity at a specified normal stress and hydraulic gradient;
• minimum strength requirements to survive installation;
• if the drainage material is a GN or core, inclusion of a GT filter above the drain; and
• if the drainage material is a GN or core, inclusion of a GT beneath the drain, if necessary,
to increase interface friction, reduce deformation-related intrusion of an underlying
DRAFT - DO NOT CITE OR QUOTE 2-49
-------
hydraulic barrier material into the drain, and/or protect the hydraulic barrier from
puncture or other damage by the drain.
As with the hydraulic conductivity of a granular drainage layer, no specific minimum hydraulic
transmissivity is recommended for a geosynthetic drainage material because the required value is
site dependent. To minimize the potential for excessive erosion and slope instability, however,
the drainage layer should be able to convey the maximum flow rate entirely in the layer without
buildup of excess head. It is noted that a geosynthetic drainage layer is generally required to
have a higher transmissivity than that for a granular drainage layer to convey the required design
flow rate under unconfined flow conditions. As discussed by Giroud et al. (2000), the
geosynthetic drainage layer hydraulic transmissivity that is equivalent to a granular drainage
layer hydraulic transmissivity for these conditions can be calculated as:
9dg = E 9ds = E kds I* (Eq. 2.19)
where: 0dg = geosynthetic drainage layer transmissivity (m3/s/m); E = equivalency factor
(dimensionless); 0ds = granular drainage layer transmissivity (mVs/m); kds = granular drainage
layer hydraulic conductivity (m/s); and tds = granular drainage layer thickness (m). The
equivalency factor can be approximated as (Giroud et al., 2000):
E =
1
0.88
1 +
cosP
0.88L
d
(Eq. 2.20)
where: Ld = length of drainage layer flow path (m), and all other terms are as defined previously.
The hydraulic transmissivity of geosynthetic drainage layers can be measured in the laboratory
using ASTM D 4716. The test setup should simulate the actual field system as closely as
possible in terms of boundary conditions, stresses, and gradient.
2.4.2.2 Thickness of Granular Layers
The recommended minimum thickness of a granular drainage layer is usually 0.3 m. This allows
sufficient thickness for ease of construction and to avoid damage to underlying geosynthetics,
such as a GM. With extremely careful control of thickness, it is possible to construct thinner
granular drainage layers (down to a thickness of about 0.15 m), but granular drainage layers
thinner than 0.3 m are not very common.
2.4.2.3 Required Flow Capacity
The flow capacity, qc (m3/s/m), of a drainage layer must be equal to or greater than the product
of the maximum flow rate, qm (m3/s/m), considered for design and the factor of safety, FS
(dimensionless):
qc>qmFS (Eq. 2.21)
As previously mentioned, the maximum flow rate can be calculated considering the cover system
water balance for the selected design storm. Methods for calculating the maximum flow rate are
presented in Section 4.5. The FS selected for design should be based on the level of uncertainty
DRAFT - DO NOT CITE OR QUOTE 2-50
-------
inherent in the design input parameters and the consequences of failure. A minimum FS value of
2 is recommended for cases where the uncertainty in input parameters is low and the
consequences of failure are small. For many situations, a larger FS may be appropriate. Koerner
and Daniel (1997) have recommended using a FS value of at least 5 to 10 to account for
uncertainities in the hydraulic conditions.
For granular drainage layers, the drainage layer hydraulic conductivity is selected to provide
adequate flow capacity and unconfined flow conditions. For geosynthetic drainage layers, the
drainage layer hydraulic transmissivity is selected to provide adequate flow capacity and
unconfined flow conditions. For all drainage layer materials, the required field hydraulic
properties for design are evaluated considering the material properties measured in the laboratory
and reduction factors that consider the potential for reduction in the property over time due to
long-term clogging, deformation, etc. in the field.
For granular drainage layers, the field hydraulic conductivity can be computed as:
kf=kj - - - (Eq. 2.22)
^RFCCRFBC J
where: kf = long-term field hydraulic conductivity of granular drainage layer (m/s); kj =
hydraulic conductivity of granular drainage layer (m/s) measured in the laboratory; RFcc =
reduction factor for chemical clogging (dimensionless); and RFec = reduction factor for
biological clogging (dimensionless).
For geosynthetic drainage layers, the field hydraulic transmissivity can be computed as:
6f = 0, - - - (Eq. 2.23)
where: 0f= long-term field hydraulic transmissivity of geosynthetic drainage layer (mVs/m); 61 =
hydraulic transmissivity of geosynthetic drainage layer (mVs/m) measured in the laboratory;
RFiN = reduction factor for elastic deformation and/or or intrusion of the adjacent geosynthetics
into the drainage layer (dimensionless); RFCR = reduction factor for creep deformation of the
drainage layer and/or creep deformation of adjacent materials into the drainage layer
(dimensionless); and all other variables are as defined previously.
It may occasionally be necessary to consider other reduction factors, such as factors for
installation damage or elevated temperature effects. If necessary, they can be included on a site-
specific basis. On the other hand, if the reduction factor has been included some way in the test
procedure for measuring the hydraulic property, the reduction factor would appear in the
foregoing formulation as a value of unity. Information on preliminary reduction factor values is
given in Koerner (1998).
DRAFT - DO NOT CITE OR QUOTE 2-51
-------
2.4.2.4 Drainage Layer Outlets
As previously discussed, water collected in a drainage layer should be conveyed to an outlet. If
there are not a sufficient number of outlets or if the outlets become clogged, the hydraulic head
in the drainage layer can build up and exceed the drainage layer thickness, leading to saturation
of cover soils and increases in seepage forces. There have been cases of significant cover system
erosion and slope instability caused by inadequate outlet design.
Drainage layer outlets are usually designed to release water into drainage ditches or swales on
the cover system or along the facility perimeter. The drainage layer may extend to the ditch or
swale, as in Figure 2-5(a) or may be connected to the drainage structure via pipes or other means.
When it is necessary to prevent the drainage layer from freezing, the drainage layer is usually
insulated with an adequate thickness of cover soil (see Section 2.3.2.2.2). However, the
prevention of freezing (and, hence, plugging) of outlet points can be challenging because outlets
are usually exposed to freezing temperatures. Pipe outlets may be more problematic than areal
outlets because they concentrate flow from a larger area. Thus, if a pipe is plugged with frozen
water, water would have to flow laterally for some distance to reach another pipe. The authors
are aware of situations where pipes plugged with ice have been dealt with as a maintenance issue
by removing the ice using a heat source.
2.4.3 Construction
The construction, quality control (QC), and CQA of granular drainage layers and the
manufacturer, installation, QC, and CQA of geosynthetic drainage layers are discussed in detail
by Daniel and Koerner (1993, 1995). This discussion is not repeated herein.
In brief, granular drainage material is usually loosely dumped from a truck and spread with a
low-ground pressure bulldozer. Low-ground pressure equipment is used to minimize the
generation of fines and the potential for damage of any underlying geosynthetics. Granular
drainage layers are generally not compacted.
Geosynthetic drainage layers are manufactured in panels of certain widths and lengths. The
panels are placed in the field and connected by overlapping, seaming, tying, interlocking, or
other means.
2.4.4 Maintenance
Maintenance is discussed in Chapter 9. Since the drainage layer is overlain by the surface and
protection layers, drainage layer maintenance is generally not needed unless the cover soils are
breached due to erosion or there are problems with excessive differential settlement or slope
instability.
2.4.5 Monitoring
Monitoring is discussed in Chapter 8. If the cover system water balance is being assessed, lateral
drainage from the drainage layer may be monitored.
DRAFT - DO NOT CITE OR QUOTE 2-52
-------
2.5 Hydraulic Barrier
The primary function of the hydraulic barrier is to limit percolation of water through the cover
system to an amount less than or equal to the maximum acceptable value. The hydraulic barrier
achieves this by impeding infiltration into the barrier and by promoting storage or lateral
drainage of water in the overlying layers. For wastes that generate gases or have volatile
constituents, the hydraulic barrier can also restrict migration of these pollutants through the cover
system and into the atmosphere.
2.5.1 General Issues
By definition, the hydraulic barrier must provide high impedance to flow of water, typically by
having a very low saturated hydraulic conductivity. The most important concern with respect to
the hydraulic barrier is the ability of the barrier to function as intended over time. Depending on
the barrier material selected, the water impedance capabilities of a barrier can become
substantially reduced when the barrier is subjected to deformations, wet-dry cycles, freeze-thaw
cycles, and biointrusion. Even when not subjected to these stresses, barriers may degrade over
time, for example, as GMs do as they lose their oxidizers by volatilization.
2.5.2 Elements of Design
Important questions that typically need to be addressed when considering the design of the
hydraulic barrier include:
• What materials are available to construct the hydraulic barrier?
• What thickness of hydraulic barrier material is needed?
• What is the expected performance of the hydraulic barrier in terms of quantity of water
percolation through the layer?
• What is the expected performance of the hydraulic barrier in terms of prevention of gas
release to the atmosphere?
• How much differential settlement is expected, what level of tensile strain will this create
in the hydraulic barrier, and how is the barrier expected to perform under this stressor?
• What is likelihood that the hydraulic barrier will be subjected to wet-dry cycles and how
is the barrier expected to perform under this stressor?
• What is likelihood that the hydraulic barrier will be subjected to freeze-thaw cycles and
how is the barrier expected to perform under this stressor?
• What hydraulic barrier properties are required to provide the required shear strength?
• What is likelihood that the hydraulic barrier will be subjected to biointrusion and how is
the barrier expected to perform under this stressor?
• What is the anticipated lifetime of the barrier material(s)?
• How should the hydraulic barrier be constructed?
• What type and frequency of maintenance should be employed?
DRAFT - DO NOT CITE OR QUOTE 2-53
-------
2.5.2.1 Materials
Materials used for hydraulic barriers include GMs, GCLs, and CCLs. Although other materials
have been used (e.g., asphaltic concrete, as discussed in Section 2.2.2.2.6), the vast majority of
all barriers are composed of one or a composite of the three materials listed above. Choices in
the composite category typically are GM/GCL, GM/CCL, or GM/GCL/CCL. It has been shown
that, all else being equal, a cover system with a composite barrier consisting of GM/CCL,
GM/GCL, or GM/GCL/CCL allows less percolation than a cover system with a GM, GCL, or
CCL barrier alone.
Each type of barrier has advantages and disadvantages. No one type should be viewed as
optimal for all cover systems. The appropriate material(s) should be selected based on the
specific objectives of a particular project and the expected site conditions.
2.5.2.1.1 GMs
GMs are thin, factory-manufactured polymeric materials that are widely used as hydraulic
barriers in cover systems due to their non-porous structure, flexibility, and ease of installation.
GMs have the advantages of extremely low rates of water and gas permeation through intact
GMs and, depending on the material, the ability to stretch and deform without tearing. They also
protect underlying CCLs from desiccation or root penetration. Disadvantages of GMs include
leakage through occasional GM imperfections, the potential for slippage along interfaces
between GMs and adjacent materials, and, for some applications, uncertainty about the length of
the GM useful service life.
GMs form an essential part of many cover system hydraulic barriers. They are manufactured in
panels, which vary in dimension depending on the manufacturing process and project-specific
criteria. The most common types of GM polymers used in cover systems include:
• HOPE;
• very flexible polyethylene (VFPE) (this classification includes linear low density
polyethylene (LLDPE), low density linear polyethylene (LDLPE), and very low density
polyethylene (VLDPE));
• flexible polypropylene (fPP);
• flexible polypropylene reinforced (fPP-R), which is fabricated with a reinforcing scrim
between two plys of polymer sheets; and
• polyvinyl chloride (PVC).
New materials are under development, and the above list of currently-used GMs should not be
viewed as a complete list of all types of GMs that might be suitable for use in a landfill cover
system. All of these GM materials are available with smooth and textured surfaces for increased
friction and, thus, shear strength when used on steep sideslopes. Additionally, spray-on
elastomeric GMs are possible, as are bituminous GMs. However, these groups are rarely used in
cover systems and, therefore, are not discussed further.
DRAFT - DO NOT CITE OR QUOTE 2-54
-------
Much has been written about the relative advantages and disadvantages of various GM materials.
It is important that the requirements of a GM for a liner system not be confused with
requirements for a cover system. In a typical liner system application, the GM is exposed to
leachate and subjected to relatively high normal stresses. Replacement or repair of the GM after
waste placement is not typically possible. Most liners are installed on firm subgrade, so the
stress-elongation characteristics of the GM are of secondary importance. The most commonly
used GM material for liner systems has historically been HDPE. Engineers have often selected
this material because of its very good chemical resistance and service life characteristics.
In cover systems, the GM is not usually exposed to leachate, although it may be exposed to
rising gases, which will often contain trace amounts of volatile constituents, or to vapors. Cover
system GMs are subjected to relatively low normal stresses. However, as cover system GMs are
often placed over compressible waste materials, which undergo post-closure differential
settlement, the stress-elongation characteristics of the GM can be an important design
consideration. While HDPE GMs have been widely used in cover systems, flexible GM barriers
made of PVC, VFPE, and fPP are finding wider use.
In the current state-of-practice, chemical compatibility is rarely considered for cover system
GMs since the upper surface of the GM is only exposed to water infiltration through the cover
soils. However, the lower surface of the GM may be exposed to gases and vapors that may
contain chemicals that are harmful to certain GM formulations. Thus, chemical resistance is an
issue that may need to be considered under site-specific conditions.
Specifications for GM hydraulic barriers often require:
• resin and additive requirements;
• limitations on the amounts of fillers, carbon black, and regrind/recycle material that can
be added to the resin;
• texture quality (e.g., minimum asperity height), if texturing is used;
• minimum thickness;
• mass per unit area; and
• minimum strength and elongation requirements.
Protection layers are often placed above a GM if angular gravel or crushed rock will be placed
on the GM. A protection GT used in this application is sometimes referred to as a cushion. In
cover systems, the overburden stresses produced by cover soils are normally not very large,
which makes the design of a GT cushion relatively simple compared to a situation in which the
angular stone overlying the GM is subjected to high compressive stresses. Procedures for
selecting a GT mass per unit area to adequately protect the GM are provided by Koerner (1998).
2.5.2.1.2 GCLs
GCLs are thin, factory-fabricated products containing a layer of sodium bentonite (a very low
permeability clay) that is supported by one or two layers of geosynthetics. GCLs have attractive
features for cover system applications, including a very low saturated hydraulic conductivity
DRAFT - DO NOT CITE OR QUOTE 2-55
-------
(e.g., typically less than 5 x 10"11 m/s, which is lower than for CCLs), preservation of low
hydraulic conductivity when subjected to different stressors, and ease of installation.
Disadvantages of GCLs include low internal shear strength of hydrated bentonite, potentially low
interface shear strength at its upper and lower surfaces (depending on the type of GCL and
interfacing materials), potential for increased hydraulic conductivity due to cation exchange
reactions under certain conditions, potential for premature hydration during installation
desiccation cracking of the bentonite layer, and root intrusion for unprotected GCLs. Although
GCLs are relatively new (first used in a waste containment application in the late 1980s), their
use has increased rapidly in the past decade. One of the more common applications of GCLs is
as the soil component of composite hydraulic barriers. Less frequently, they are used alone as a
barrier. The results of a large-scale field test program sponsored by EPA to evaluate GCL use in
cover systems are summarized in Section 7.4.5.
GCLs consist of sodium bentonite placed between GTs and mechanically held together by
adhesive or fibers, or bentonite adhesively bonded to a GM or GT/GM laminate. The types of
GCLs most commonly used in cover system applications are shown in Figure 2-12. The
bentonite is the low-hydraulic conductivity component; the geosynthetics act as carrier materials
or, in the case of GCLs incorporating GMs, as a supplemental hydraulic barrier. The carrier
geosynthetics support the bentonite component and help to maintain a uniform layer of bentonite
that can be handled, transported, and placed as a barrier. The manufactured material has a
nominal clay thickness of 5 mm and is produced on rolls that measure about 4 m in width and 30
to 60 m in length. The mass of bentonite per unit area (dry weight basis) is typically at least 3.6
kg/m2.
Bentonite is the critical component of GCLs. Bentonite is a naturally occurring, mined clay
mineral material that is extremely hydrophilic. When placed in the vicinity of water (or even
water vapor), the bentonite attracts water molecules into a complex configuration that leaves
little free water space in the voids. This significantly decreases the hydraulic conductivity of the
bentonite. When the bentonite is saturated and permeated with fresh water, the hydraulic
conductivity is typically on the order of 1 to 5 x 10"11 m/s, or less, depending on the bentonite
and the effective confining stress used in the measurement of hydraulic conductivity. Because
hydraulic conductivity decreases with increasing effective confining stress, it is important that
the effective confining stress be reported along with hydraulic conductivity. For cover system
applications, it is common to report hydraulic conductivity at an effective confining stress of
approximately 35 kPa, which is the lower limit of effective confining stress that is recommended
for routine commercial hydraulic conductivity testing of GCLs.
GCLs can be reinforced by needlepunched fibers or stitching that increases the internal shear
strength of the GCL, which can help to maintain stable slopes. A variety of woven and
nonwoven GTs can be used. For GM-supported GCLs, the GM can be smooth or textured, and
the thickness can be as little as 0.3 mm or as much as 2 mm. New types of GCLs are being
developed, and the materials and configurations are continually expanding and improving.
DRAFT - DO NOT CITE OR QUOTE 2-56
-------
Woven or Nonwoven GT
Needlepunched
Fibers-
Sc
dil
jm
Be
snt
Dni
te
^
(a)
Nonwoven GT
(b)
Woven GT
Sewn Stitche
\
SodiurmBentonite
Mixed with! an Adhesive
i 1
Woven GT
~7
(c)
Sodium Bentonite
Mixed with an Adhesive
GM
Figure 2-12. Types of GCLs Commonly Used as Cover System Barriers: (a) Reinforced,
GT-Encased, Needlepunched GCL; (b) Reinforced, GT-Encased, Stitch-
Bonded GCL; and (c) Unreinforced, GM-Supported GCL.
Specifications for GCL hydraulic barriers often require:
• restrictions on bentonite properties (minimum free swell, maximum fluid loss);
• minimum mass per unit area;
• minimum strength and strain requirements; and
• maximum hydraulic conductivity.
Three EPA reports on GCLs have been published (Daniel and Estornell, 1991; Daniel and
Boardman, 1993; and Daniel and Scranton, 1996). A detailed discussion of GCLs is provided by
Koerner(1998).
DRAFT - DO NOT CITE OR QUOTE
2-57
-------
In GCL applications, it is important to ensure that the hydraulic conductivity of the GCL is not
adversely affected by post-installation chemical changes. The bentonites used in GCLs are
sodium-based, which means that the dominant exchangeable cation in the pore water of the
bentonite is sodium. When GCLs are placed in contact with soils, the bentonite in the GCL
begins to absorb water immediately from the adjacent soils, unless a GM separates the bentonite
from the adjacent material. The hydration process is relatively rapid, with significant hydration
occurring in a few days and nearly complete hydration occurring within a few weeks. If the
cations in the hydrating liquid contain a mix of monovalent and polyvalent cations, little
alteration in hydraulic conductivity normally occurs. However, if the hydrating water is rich in
polyvalent cations such as calcium, the GCL may not swell adequately or attain the desired low
hydraulic conductivity. Even if a GCL is initially hydrated with a water containing few
polyvalent cations, the GCL may be affected in the long term if it is permeated by an infiltrating
water rich in polyvalent cations. Over time, the indigenous sodium cations in the GCL may be
replaced by the polyvalent cations. Calcium-rich soils, or aggregates containing limestone, are
of particular concern because they leach calcium. Melchior (1997a,b) and James et al. (1997)
document cases in which cation exchange converted the sodium bentonite in GCLs used for
cover systems to calcium bentonites, causing an increase in hydraulic conductivity. If the
potential exists for teachable cations in overlying surface, protection, or drainage layers to
adversely impact GCL hydraulic conductivity, this impact should be evaluated by index testing
(e.g., free swell and fluid loss tests) and by hydraulic conductivity testing, for example, as
described by Ruhl and Daniel (1997). If necessary, the GCL should be protected with a GM or
different materials should be used above the GCL.
One of the potential problems with GCLs is thinning of bentonite if the GCL is placed on sharp
objects such as stones or sharp changes in local topography, such as ruts left by vehicles. To
avoid these problems, it is recommended that no protruding stones larger than approximately 12
mm be present on the subgrade surface, and that no ruts deeper than about 25 mm be present.
GCLs need to be covered with a GM or an adequate thickness of soil as soon as possible after
installation to prevent unconfmed hydration. If the GCL hydrates while unloaded, the GCL can
swell excessively and potentially extrude laterally as overburden soil is placed. The hydrated
GCL also has relatively low shear strength and may impact slope stability. Even if the GCL is
covered with a GM, there is still potential for hydration if the underlying subgrade materials are
wet or if the waste emits gases that are saturated with water vapor. Daniel et al. (1993) and
Bonaparte et al. (1996) provide data on GCL hydration due to contact with compacted subgrade
soil.
GCLs also need to be covered with an adequate thickness of soil prior to operating heavy
vehicles above the GCL. If adequate protection is not provided, the bentonite can extrude
laterally, causing localized thinning (Koerner and Narejo, 1995). Experience from tests reported
by Koerner and Narejo (1995) and Fox (1998) indicates that bentonite will not be squeezed
laterally in the GCL as long as the thickness of cover soil is at least one to two times greater than
the width of the tire load at the surface of the protective soil layer. Based on this, the minimum
thickness of cover soil should be about 0.45 to 0.6 m. This should be accomplished in practice
since at least 0.3 m of soil is generally maintained between geosynthetics and low-ground
DRAFT - DO NOT CITE OR QUOTE 2-58
-------
pressure tracked equipment and at least 0.6 m of soil is generally maintained between
geosynthetics and rubber-tired vehicles.
2.5.2.1.3 CCLs
CCLs are constructed from materials that are mineralogically stable and are well known to
design engineers, regulators, and contractors. CCLs offer the advantage over GMs and GCLs in
that they are much thicker, which makes them much less susceptible to accidental puncture.
Historically, CCLs have been the most frequently used cover system barrier material.
Procedures for construction of CCLs to meet permeability criteria are well-established.
However, information developed more recently indicated that, when used alone, CCLs in cover
systems may not maintain their low permeability in the long term. This is particularly true if a
CCL hydraulic barrier is used at an arid or semi-arid site, is located above the depth of frost
penetration, or has insufficient overlying cover soil to prevent desiccation cracking. Section 7.2
summarizes a number of field case histories where CCL barriers in cover system applications
exhibited increasing permeability with time, even when the CCLs were overlain by cover soils.
The increase in permeability is attributed to wet-dry and freeze-thaw effects, root penetration,
and differential settlement. Bonaparte et al. (2002) suggest that the best way to maintain low
CCL permeability in this application is to overlay the CCL with both a GM and a cover soil with
a thickness sufficient for the site-specific conditions. Another limitation of CCLs is their
inability to conform to all but the smallest differential settlements of the underlying waste
without cracking. Tension cracks starting from the underside of the CCL and propagating
upwards through the thickness of the CCL can render them nearly useless as barriers to water
infiltration or gas release.
CCLs are constructed primarily from natural soil materials that are rich in clay, although the
barrier may also contain processed materials such as bentonite. Specifications for CCLs that
must have a hydraulic conductivity of not more than 1 x 10"9 m/s often require (Koerner and
Daniel, 1997):
• minimum percentage of fines (particles passing the No. 200 sieve (0.074 mm openings))
> 30-50%;
• minimum plasticity index > 7-15%;
• maximum percentage of gravel (particles retained on the No. 4 sieve (4.76 mm openings)
< 20-50%; and
• maximum particle size < 25-50 mm (perhaps less for lifts overlain by a GM).
Local experience may dictate different requirements, and, for some soils, more restrictive criteria
may be appropriate. However, if the criteria tabulated above are not met, it is unlikely that a
natural soil liner material will be suitable without additives such as sodium bentonite.
If there is concern that rocks or stone in the CCL material may damage an overlying GM, the
stones should be removed. Vibratory screens can be used to sieve stones prior to placement or
mechanical devices that remove stones in a loose lift can be used. A different material, or a
differentially processed material that has fewer and smaller stones, may also be used to construct
the uppermost lift of the CCL to be covered by a GM.
DRAFT - DO NOT CITE OR QUOTE 2-59
-------
CCLs used in cover systems should be as ductile as possible (to accommodate differential
settlement) and should be resistant to cracking from moisture variations (e.g., desiccation).
Sand-clay mixtures are ideal materials if resistance to shrinkage and desiccation-induced
cracking are important (Daniel and Wu, 1993). Ductility is achieved by avoiding use of dense,
dry soils that tend to be brittle. If suitable materials are unavailable, local soils can be blended
with commercial clays (e.g., bentonite) to achieve low hydraulic conductivity. A relatively small
amount of sodium bentonite (typically 2 to 6% by weight) can lower hydraulic conductivity as
much as several orders of magnitude. The percent bentonite is usually defined as the weight of
bentonite (including a small amount of hydroscopic water) divided by the weight of soil (dry and
moist weight have been used, but the dry weight is recommended) to which bentonite is added.
Soils with a broad range of grain sizes usually require a relatively small amount of bentonite
(i.e., less than 6%). Uniform-sized soils, such as dune sand, usually require more bentonite (i.e.,
up to 10-15%). Sometimes different soils are blended to provide a material with a broad range of
grain sizes, thus reducing the amount of bentonite needed to achieve the specified hydraulic
conductivity criterion. For instance, on one project, a coarse to medium sand was successfully
blended with bentonite (Alston et al., 1997). By adding 30% of fine, inert material (waste fines
from a materials processing plant), the amount of bentonite required was halved. In some cases,
GCLs are selected over soil-bentonite CCLs due to economics or ease-of-construction
considerations.
2.5.2.2 Thickness
2.5.2.2.1 GMs
The thickness of a GM used in a cover system is selected based upon several factors, the most
important of which are durability and capability of being seamed. GMs should be adequately
thick to resist construction damage and puncture. The minimum recommended thickness for this
purpose is thought to be 0.75 mm. The minimum thickness for adequate field seaming varies
with material but is typically in the range of 0.75 to 1 mm. As the GM thickness increases, other
mechanical properties also increase. Koerner (1998) suggests that the GM properties given in
Table 2-7 be used as a guide to installation survivability, i.e., the ability to be installed without
significant damage. GMs should be selected with sufficient thickness to meet the material
properties in this table.
2.5.2.2.2 GCLs
GCLs are manufactured with a nominal clay thickness of 5 mm. Like GMs, GCLs are thin and
may potentially be punctured during installation. Unlike GMs, however, GCLs possess
significant self-sealing capability due to the swelling of dry bentonite upon hydration or the
plastic flow of hydrated bentonite. Shan and Daniel (1991) found that holes as large as 25 mm in
diameter in a dry GCL swelled shut when the GCL was hydrated, and that the hydraulic
conductivity was not significantly affected by the large puncture. However, it is possible to
puncture GCLs (e.g., with construction equipment) to the point that self-sealing will not occur.
DRAFT - DO NOT CITE OR QUOTE 2-60
-------
Table 2-7. Minimum properties for general GM installation survivability suggested by
Koerner(1998).
Property and Test Method
Thickness (ASTM D 1593) (mm)
Tensile (ASTM D 682, 25 mm strip) (kN/m)
Tear (ASTM D 1004 Die C) (N)
Puncture (ASTM D 4833) (N)
Impact (ASTM D 3998 mod.) (J)
Required Degree of Installation Survivability1
Low
0.63
7.0
33
110
10
Medium
0.75
9.0
45
140
12
High
0.88
11
67
170
15
Very High
1.00
13
90
200
20
1 Low refers to careful hand placement on a uniform, well-graded, smooth subgrade with light loads of a static nature,
typical of vapor barriers beneath building floor slabs.
Medium refers to hand or machine placement on a machine-graded subgrade with medium loads, typical of canal
liners.
High refers to hand or machine placement on a machine-graded subgrade of rough texture with high loads, typical of
landfill liner and cover systems.
Very high refers to hand or machine placement on machine-graded subgrade of very rough texture with high loads,
typical of liners for heap leach pads and floating covers for impoundments.
2.5.2.2.3 CCLs
CCLs are constructed in layers called "lifts" that typically have a thickness before compaction
("loose lift") of 0.2 to 0.25 m and a thickness after compaction ("compacted lift") of not more
than 0.15 m. Typically three to six lifts are used to produce a CCL hydraulic barrier with a final
thickness of 0.45 to 0.9 m. Since each lift of CCL may potentially have areas that do not meet
the hydraulic conductivity criterion (as construction of CCLs is, by nature, less controlled than
the manufacture of GMs and GCLs), the use of multiple lifts decreases the likelihood that these
areas would be continuous through the CCL thickness. A minimum of three compacted lifts is
recommended. If the CCL hydraulic barrier is not overlain by a GM, four of more compacted
lifts is preferred. It is noted that these recommendations on minimum CCL thickness are based
on constructability and performance considerations, not minimum regulatory guidance, which in
some cases may allow a thinner CCL.
2.5.2.3 Percolation
The selection of the hydraulic barrier depends to some extent on the allowable rate of water
percolation through the cover system. In most instances, the cover system is intended to allow
very little infiltration of water into the waste, and the hydraulic barrier is essential to achieving
low percolation rates. In other instances, particularly those involving risk-based corrective
actions, the amounts of percolation may be less restrictive.
It is recommended that the percolation objective for the cover system be defined, at least
qualitatively, prior to design. Methods for estimating percolation rates through cover systems
are presented in Chapter 4.
DRAFT - DO NOT CITE OR QUOTE
2-61
-------
Liquids can migrate through GMs by two mechanisms: (i) permeation through an intact GM; and
(ii) flow through GM holes. Fluids permeate GMs by molecular diffusion. The process involves
adsorption of the diffusing chemical or compound into the surface of the GM, diffusion through
the GM, and desorption from the opposite surface of the GM. Some diffusion rates reported in
the literature for GMs are as follows:
• 1.0 mm-thick HDPE: water vapor transmission (WVT) rate = 0.020 g/m /day;
1.0 mm-thick HDPE: solvent \
depends on solvent type); and
• 1.0 mm-thick HDPE: solvent vapor transmission (SVT) rate = 0.02 to 20 g/m2/day
• 0.75 mm-thick PVC: WVT rate =1.8 g/m2/day.
The WVT values are relevant for infiltrating water coming through the cover soil and eventually
entering into the underlying waste mass. The SVT values are relevant if there are rising vapors
or gases from the waste mass. For MSW landfills, the gases are saturated with water vapor and
may contain low concentration of solvents derived from volatilization within the landfill.
Diffusion coefficients for various organic solvents and polyethylene GMs are summarized by
Rowe (1998). The above WVT rates are relatively low and do not result in significant amounts
of water percolation through the hydraulic barrier. While the SVT rates are higher, solvent mass
transfer through GM hydraulic barriers will, in most cases, be very low due to the low
concentration of solvents in any gas in contact with the barrier layer. The authors caution,
however, that while solvent mass transfer through the cover system will be insignificant in most
cases, it should be considered in evaluating GM barriers used for capping of remediation source
areas which may contain a significant solvent mass.
Of greater significance than water vapor diffusion is flow through GM holes, such as tears,
punctures, or imperfect seams. Flow through such holes in a GM alone usually significantly
exceeds the diffusion values listed above (EPA, 1991). If the GM is underlain by a GCL or CCL
to form a composite barrier, water migrating through a GM hole or defect will be impeded by the
underlying GCL or CCL. Flow through the GCL or CCL will then be limited by the area of the
GM hole(s), which is only a small fraction of the total area of the barrier, and any lateral flow at
the interface of the GM and the GCL or CCL. The amount of interface flow is a function of the
"intimacy" of the contact between the GM and GCL or CCL components (Giroud and
Bonaparte, 1989b; Gross, et al., 1990). If there is good contact between the GM and underlying
GCL or CCL, the flow rate through a GM hole will be very low (unless the hydraulic head acting
on the hole becomes very large, which is usually not the case). The relative performance of GM
and composite barriers is apparent when analyzing field data on apparent leakage rates through
the top liners of double-lined landfills. As described by Gross et al. (1997) and Othman et al.
(2002), the data indicate that GM barriers have a representative hydraulic efficiency of 99% and
GM/GCL and GM/CCL composite barriers have a representative efficiency of 99.9%, where
efficiency is defined as the percentage of lateral drainage that flows from the drainage layer
rather than percolates through the barrier. Methods of estimating leakage though holes in GMs
alone and GM/CCL and GM/GCL composite barriers have been presented by Giroud and
Bonaparte (1989a, 1989b), Giroud et al. (1989), Giroud et al. (1992), Giroud (1997), Rowe
(1998), and Foose et al. (2001). Recommendations on the use of the different leakage models
are presented by Foose et al. (2001).
DRAFT - DO NOT CITE OR QUOTE 2-62
-------
Percolation through GCL or CCL barriers is typically estimated using Darcy's equation for
saturated conditions or Richards' partial differential equation for unsaturated conditions
(Richards, 1931).
2.5.2.4 Gas Containment
When there is a need for gas containment, GMs are generally the best barriers to gas. GCLs and
CCLs also make very good gas barriers when they are at high degrees of saturation and do not
contain major secondary structures, such as desiccation cracks extending through the GCL or
CCL.
2.5.2.5 Differential Settlement
Differential settlement is usually quantified in terms of the magnitude of differential settlement
(A) that occurs over a distance (b), yielding angular distortion, A/b (Gilbert and Murphy, 1987),
as shown in Figure 2-13. Angular distortion may damage barriers because distortion produces
tensile strains, and tensile strains can cause barrier materials to fail if the strains are excessive.
Tensile strains are generated by the material elongation associated with geometric distortion. A
relationship between angular distortion and tensile strain is shown in Figure 2-13.
Distortion, A/b
Figure 2-13. Theoretical Relationship Between Tensile Strain and Angular Distortion
(modified from Gilbert and Murphy, 1987).
DRAFT - DO NOT CITE OR QUOTE
2-63
-------
Procedures for estimating total and differential settlements are discussed in Chapter 6.4.
Frequently, the estimates of A/b that are used for design are based primarily on experience and
observations. The magnitude of A/b that is expected is highly site dependent and is a function of
variables such as type of waste, age of waste, details of waste placement, and thickness of the
cover system. The impact of settlements on hydraulic barriers is discussed in detail in Section
6.5.
The selected barrier materials should be able to accommodate the anticipated settlements. Axi-
symmetric, out-of-plane tests on various GMs have resulted in the stress-strain curves shown in
Figure 2-14. The ability of the different GMs to accommodate differential settlement is lowest
for chlorosulfonated polyethylene-reinforced (CSPE-R) and FtDPE and highest for VLDPE,
LLDPE, and PVC. As previously mentioned, VLDPE and LLDPE are both in the VFPE
category. If significant differential settlement is anticipated, as with cover system barriers over
MSW, the use of GMs that can accommodate high out-of-plane, or axisymmetric, deformations
should be considered.
Test results published by Koerner et al. (1996) and LaGatta et al. (1997) indicate that reinforced
GCLs can withstand tensile strains of 5 to 16%, depending on product. Care should be taken to
ensure an adequate overlap width, since, under elongating conditions, slippage may occur along
overlaps.
CCLs can accommodate little tensile elongation. As described in Section 6.5, CCLs will
typically exhibit tensile failure at extensional strains of 0.5% or less.
ro
CL
CO
HOPE
VLDPE
LLDPE
PVC
CSPE-R
40 60
Tensile Strain (%
100
Figure 2-14. Stress-Strain Behavior of Common GM Materials Subjected to Axi-
Symmetric, Out-of-Plane Tensile Strain (modified from Koerner et. al., 1990).
DRAFT - DO NOT CITE OR QUOTE
2-64
-------
2.5.2.6 Wet-Dry Cycles
The potential for wet-dry cycles to affect the integrity of CCLs and, to a lesser extent, GCLs,
should be considered whenever these materials are used as hydraulic barriers. Water balance
analyses, such as those described in Chapter 4, can be helpful, but judgment should play an
important role in the evaluation process. If damage to a CCL or GCL is anticipated, the normal
solution is to use a composite GM/CCL or GM/GCL hydraulic barrier overlain by a protection
layer.
Cyclic wetting and drying can have a significant impact on the hydraulic conductivity of CCLs
under low confining pressures. As drying progresses, shrinkage occurs and reaches a limit at
which cracking can occur. This cracking, caused by desiccation, occurs in block form, and
gradually progresses deeper into the CCL until a pathway of water migration becomes available.
Besides drying as a result of ET, CCLs may also lose moisture to materials (e.g., a dry soil
foundation layer) beneath them.
Both soil dry density and soil water content affect the vulnerability of the soil to desiccation
cracking (Albrecht and Benson, 2001). Highly plastic clays undergo large shrinkage when dried;
clayey sands undergo little shrinkage. A given CCL material experiences less shrinkage when it
is compacted at its optimum moisture content and with a high compactive effort as compared to
the shrinkage of the same soil compacted to wetter or less dense conditions. Shrinkage and
cracking can occur in CCLs as a result of water content changes of only 2 to 5 percentage points.
Moisture content variations of this magnitude are inevitable in the top 1 to 2 m of soil at most
sites. With the reintroduction of water, swelling occurs and the cracks start to close. However,
the degree to which the cracks swell shut is highly dependent on overburden pressure (Boynton
and Daniel, 1985). At overburden stresses of less than 40 to 100 kPa, cracks do not fully close,
even after the soil is soaked. The overburden stress on CCLs in cover systems is typically less
than 25 kPa. Thus, in cover systems, the remnants of desiccation cracks are likely to remain,
causing the hydraulic conductivity to increase over its as-constructed value.
Experience has shown that severe desiccation can occur to depths of up to 1 m, and possibly
deeper (Montgomery and Parsons, 1989, 1990; Corser and Cranston, 1991; Corser et al., 1992;
Melchior et al., 1994; Melchior, 1997a,b; Maine Bureau of Remediation and Waste
Management, 1997; and Khire et al., 1997, 1999). The information that is available on
desiccation spans a period of field observation of approximately five years. Over longer periods,
the depth of impacts associated with wet-dry cycling could extend even deeper. It is
recommended that at least 1.2 m of cover soil, and possibly more, be used to protect the CCL
(assuming that it is not overlain by a GM) from desiccation cracking. Even greater thicknesses
(e.g., 1.5 m) may be necessary in certain cases.
Depending on the chemistry of the permeating water, GCLs may or may not be vulnerable to
permanent damage from desiccation. When permeated with water containing little salts, GCLs
are less vulnerable than CCLs to permanent damage from desiccation, because of the swelling
and self-healing capability of bentonite (Boardman and Daniel, 1996; Lin and Benson, 2000).
Data published by Shan and Daniel (1991), Boardman and Daniel (1996), and Lin and Benson
(2000) indicate that, under this condition, GCLs can withstand at least five cycles of wetting and
drying without a significant increase in long-term hydraulic conductivity. However, if the
DRAFT - DO NOT CITE OR QUOTE 2-65
-------
permeant contains cations that may exchange with the sodium in the GCL bentonite, the barrier
will loose some capability to swell and recover from desiccation over time. GCLs have been
damaged for this reason in at least several field installations (Melchior, 1997; James et al., 1997).
Though GCLs may have significant swelling and self-healing capability following wet-dry
cycles, it is not recommended that these barriers be exposed to these cycles. There is concern
that the GCLs may lose their self-healing capability over time due to cation exchange. This is
especially a concern at sites in semi-arid and arid climates, since barriers may become saturated
in the winter months and very dry in the summer months. Pore water in these environments also
tends to have higher salt concentrations than that in more humid climates.
The best approach for protection of a CCL or GCL from desiccation is to place a GM over the
barrier, and then cover the GM with soil.
2.5.2.7 Freeze- Thaw Cycles
The potential for freeze-thaw of the hydraulic barrier should be evaluated, as discussed in
Section 2.3.2.2.2. If the hydraulic barrier is located below the maximum depth of frost
penetration, then the barrier is usually assumed to be adequately protected from long-term frost
damage. If the barrier is within the zone of frost penetration, then the impacts of frost upon the
barrier materials should be considered.
Frost is generally believed to have no effect on GMs (Comer et al., 1995). This is only true,
however, if the GM is buried such that stresses induced by thermal contraction do not cause
tensile failure of a GM. An exposed GM (i.e., an exposed GM cover system) will undergo much
larger temperature fluctuations than one buried beneath a thick layer of cover soil.
Laboratory data (Hewitt and Daniel, 1997) as well as field data (Erickson et al., 1994; Kraus et
al., 1997) suggest that GCLs can withstand multiple cycles of freeze-thaw with little or no
adverse effect on the thawed hydraulic conductivity of the GCL. However, the GCL test data
available at this time are relatively short-term. In addition, there is the potential for GCLs to
become damaged if they desiccate under freezing conditions and then rehydrate with water
containing exchangeable cations. If desiccation/rehydration of GCLs is a concern, suitable
approaches for GCL protection are to place the GCL beneath a sufficiently thick soil layer or to
cover the GCL with a GM.
Freezing temperatures can cause desiccation and freeze-thaw cracking in CCLs, resulting in
barriers with increased permeability to water and gas. Desiccation cracking occurs as water is
drawn from a CCL and towards a freezing front. Freeze-thaw cracking occurs as the ice lenses
form in the CCL. Available information indicates that CCLs will not maintain a hydraulic
conductivity of 1 x 10"9 m/s or less if subjected to freeze-thaw at the level of overburden stress
normally encountered in cover systems. Instead, the CCL hydraulic conductivity will increase
by one to two orders of magnitude (Othman et al., 1994). The exception to this appears to be for
compacted soil-bentonite CCLs (Wong and Haug, 1991), which do not appear to be vulnerable to
damage from freeze-thaw action. If CCL damage by frost action is a concern, suitable
approaches for CCL protection are to place the barrier beneath a sufficiently thick soil layer or to
cover the CCL with a GM and then a soil layer.
DRAFT - DO NOT CITE OR QUOTE 2-66
-------
2.5.2.8 Shear Strength
Measurement of the shear strength parameters of different barrier materials is discussed in some
detail in Section 6.2.4. Specific issues relevant to barrier strength are discussed in this section.
GMs can have a low interface shear strength when placed adjacent to certain materials, such as
GNs or GTs. For some interfaces (e.g., GM/GT), the shear strength can be significantly
enhanced by using a textured GM. There are a number of manufacturing methods available to
provide such texturing:
• co-extrusion for blown film manufacturing;
• impingement for flat die manufacturing;
• lamination for flat die manufacturing; and
• structuring via a heated calendar for flat die manufacturing.
Perhaps the single most important design issue for GCLs that are placed in cover systems is
slope stability. When GCLs are installed on slopes, instability can occur by at least four different
mechanisms: (1) slippage at the interface between the upper surface of the GCL and overlying
material; (2) shearing within the GCL; (3) slippage at the interface between the lower surface of
the GCL and the underlying material; and (4) a combination of the first three mechanisms. The
first and third mechanisms are termed "interface" failures, and the second one is termed an
"internal" failure. Laboratory test methods to evaluate the shear strength of GCLs are discussed
in Section 6.2.4. Specific testing issues for GCLs are discussed below.
The response of GCLs to shearing stresses depends on the hydration conditions. Wet bentonite
is far weaker than dry bentonite and, therefore, the internal shear strength of hydrated GCLs can
be much lower than that of dry GCLs. An example is shown in Figure 2-15 for an unreinforced
GCL. If the GCL is expected to become hydrated by absorbing moisture from subgrade soils or
by other mechanisms, the shearing tests are normally performed on hydrated GCLs. It is
important to realize that the bentonite does not have to be completely saturated to be weakened
from hydration; the bentonite need only absorb significant moisture from the subgrade soil to
have the low shear strength of hydrated bentonite (Figure 2-16).
Reinforcement can significantly increase the internal shear strength of GCLs. As shown in
Figure 2-15, the peak failure envelope for internal shear of reinforced GCLs is much higher than
the peak failure envelope for unreinforced GCLs, but the residual strengths for reinforced and
unreinforced GCLs are about the same because at residual conditions, the internal reinforcement
has been broken.
Slippage may occur at the interface between a GCL and adjacent materials. Because GCLs may
be manufactured from woven or nonwoven GTs, and from smooth or textured GMs, a wide
range of interface shear responses may be observed. Further, GCLs may interface with a wide
range of soil and geosynthetic materials. No general statements can be made about the actual
shear strength of interfaces: there are so many permutations possible that each specific interface
should be evaluated through interface shear testing.
DRAFT - DO NOT CITE OR QUOTE 2-67
-------
Normal Stress
Figure 2-15. Comparison of Shear Strengths for Internally Reinforced GCLs and
Unreinforced GCLs.
50 75 100
Bentonite Water Content (%)
125
150
Figure 2-16. Effect of Bentonite Water Content on Shear Strength of an Unreinforced
GCL (modified from Daniel et al, 1993).
DRAFT - DO NOT CITE OR QUOTE
2-68
-------
Experience has shown that certain design situations involving GCLs installed on slopes warrant
particular attention:
• Cover system (i.e., low normal stress and no seepage forces) slopes that are inclined at
6H: IV or flatter will be stable with a FS of 1.5 or more with respect to unreinforced GCL
internal shear strength and interfaces with the GCL. Steeper slopes may also be stable
but require careful testing and analysis.
• Those GCLs with woven slit-film GTs on one or both surfaces should be carefully
evaluated to be sure that hydrated bentonite does not extrude and lubricate the adjacent
material interface (upper and/or lower) and cause a reduction in interface shear strength
compared to the shear strength in the absence of extrusion.
• Designs that rely on the dry shear strength of GCLs for stability should assure that the
GCLs will be fully and completely protected against hydration. This is usually possible
only by having GMs on both surfaces of the GCL and, in addition, having construction
and deployment conditions in the field that do not allow the GCL to absorb moisture.
• The internal shear strengths of needlepunched and stitch-bonded GCLs appear to be
adequate to achieve internal stability of the GCLs on cover system slopes as steep as
2H: IV with a FS of 1.5 or more. However, interface shear strengths for these types of
GCLs at cover system normal stresses will often be less than the internal shear strength
and at a 2H: IV slope it is likely that the cover system will be unstable or only marginally
stable.
• For cover systems with soils and textured GMs having interfaces with internally-
reinforced GCLs, slopes as steep as 3H: IV can be constructed and remain stable at a FS
of 1.5 or more (in the absence of seepage forces), but actual stability depends on the
particular materials used.
• Woven GTs generally have lower interface shear strength with materials such as soil or
other geosynthetics than non-woven GTs. If high interface shear strength is required with
a GT-encased GCL, a GCL with non-woven GTs on both surfaces is usually required.
Many times the critical interface will be between a GCL and overlying GM. In this
situation, high interface shear strength is usually achieved by installing a nonwoven GT
component of the GCL with a textured GM. The fibers of the non-woven GT become
entangled with the ridges on the textured GM, creating what some have described as the
"Velcro effect" in which high adhesion is developed. However, under large deformations
along the interface, a polishing of the materials may occur, and the residual strength may
be much lower than the peak strength. Clearly, the shearing response of GCL interfaces
can be very complex and requires careful testing and engineering.
The shear strength of a CCL, and particularly a GM/CCL interface, can be critical to the stability
of a cover system. Low hydraulic conductivity is most easily achieved by adding water to the
clay and compacting it wet of its optimum water content. However, the conditions that tend to
result in a low CCL hydraulic conductivity also tend to cause low interface shear strength. The
selection of appropriate water content-density parameters is usually a compromise between the
need for low hydraulic conductivity and the need for adequate shear strength. The design
engineer should not focus solely on achieving low CCL hydraulic conductivity to the extent that
DRAFT - DO NOT CITE OR QUOTE 2-69
-------
inadequate attention is given to the shear strength of the CCL and CCL interfaces with other
materials.
2.5.2.9 Accidental or Intentional Puncture
The potential for accidental (due to construction and operational activities) or intentional breach
of the hydraulic barrier should be considered in the design of cover systems. With respect to this
issue, the thinness of both GMs and GCLs is a disadvantage in contrast to the typical thickness of
CCLs. In evaluating GCLs, however, the sealing potential of bentonite should be considered.
This is not the case for GMs. Thus CCL, GM/GCL, or GM/CCL hydraulic barriers are superior
to GM barriers alone from the standpoint of resistance to puncture.
2.5.2.10 Anticipated Lifetime
The anticipated lifetime of the barrier material should be considered in relation to the required
design lifetime of the cover system. In this regard, reference should be made to Section 1.2.6 of
this document, where a distinction is made between the minimum post-closure period and the
design life goal of a cover system. The anticipated lifetimes of the different hydraulic barrier
materials are discussed below.
2.5.2.10.1 GMs
For GMs, aging involves a gradual transition from a ductile material to a brittle material. As
embrittlement occurs, the GM does not disappear; rather settlement, deformation, seismic
vibration, etc. can cause a brittle cracking, signifying the end of the material's functional life.
The service life of any GM component of the cover system is dependent on the specific material
used and how well the material is protected. While the degradation mechanisms leading to GM
embrittlement are many, the most severe ones are eliminated by the timely protection of the GM
after installation with cover soil or other materials. For example, the potential for polymer
degradation by ultraviolet light and elevated temperature is essentially eliminated by placement
of cover soil over the GM. Furthermore, the potential for chemical degradation of a cover
system hydraulic barrier may not be an issue since the cover system is located above the waste.
The possible exception to this is for wastes that generate gases or vapors that may bring volatile
chemicals at high enough concentrations to the underside of the GM. The primary mechanism of
degradation of a GM hydraulic barrier in a cover system is oxidation of the polymers causing
embrittlement over a long time period.
Conceptually, the oxidation of GMs can be considered in three distinct stages. These stages are
designated as: (i) depletion time of antioxidants; (ii) induction time to the onset of polymer
degradation; and (iii) degradation of the polymer to decrease some properties to a defined level
(e.g., 50% of its original value). The purpose of antioxidants in a GM formulation is to prevent
polymer degradation during processing and to prevent polymer oxidation reactions from taking
place during the first stage of service life. However, there is only a limited amount of
antioxidant in any formulation. Hence, the lifetime for this stage is limited to the specific
amount of antioxidant used. Once the antioxidant is depleted, oxygen or other strong oxidizing
agents will begin to attack the polymer, leading to the induction time stage and subsequently to
the degradation of performance properties. The duration of the antioxidant depletion stage also
DRAFT - DO NOT CITE OR QUOTE 2-70
-------
depends on the type of selected antioxidant. Many different antioxidants are commercially
available, and depletion time will vary from formulation to formulation. Proper selection of
antioxidants is known to contribute greatly to the overall lifetime of the GM. For example,
Hsuan and Koerner (1996) reported an antioxidant depletion time of about 130 years at 25°C for
an HDPE GM formulation with approximate 0.5% antioxidant package. The testing was
conducted for a simulated landfill environment with the GM placed on a layer of dry sand,
covered with sand and then 0.3 m of water, and subjected to a compressive stress of 260 kPa.
Note that this antioxidant depletion time is for HDPE, which is considered to be the most stable
of polymers being used in GMs. Research is ongoing for GMs using time-temperature
superposition procedures followed by Arrhenius modeling (Hsuan and Koerner, 1998; Hsuan and
Koerner, 2002). The most extensive service life data currently available are for HDPE GMs.
Hsuan and Koerner are currently evaluating the antioxidant depletion time for other polymers in
a like manner.
In properly formulated GMs, oxidation does not begin to occur until after the depletion of the
antioxidant. Oxidation of the polymer occurs only very slowly in a buried soil environment.
The initial stage of oxygen absorption is called the induction stage. It is the time period in which
there is no measurable change in the physical-mechanical properties of the GM. The reason for
this is related to the concentration of hydroperoxide, as described below. The first step of
oxidation (after depletion of the antioxidants) is the formation of free radicals. The free radicals
subsequently react with oxygen and start chain reactions. The free radicals are highly reactive in
that they cause chain scission of the polymer backbone, which gradually results in the
embrittlement of the material. In the induction stage, little hydroperoxide is present and, when
formed, it does not decompose. As a result, accelerated oxidation reactions do not occur. As
oxidation propagates slowly, additional hydroperoxide molecules are formed. Once the
concentration of hydroperoxide reaches a critical level, decomposition of the hydroperoxide
begins and accelerated chain reactions start. This signifies the end of the induction period
(Rapoport and Zaikov, 1986). This also indicates that the concentration of hydroperoxide has a
major effect on the duration of the induction period.
The duration of the induction stage for HDPE can be estimated from data for plastic pipes and
testing conducted on HDPE waste exhumed from a landfill (Hsuan and Koerner, 2002). Viebke
et al. (1994) presented aging data for unstabilized medium density polyethylene pipes that were
tested with pressurized water inside and circulating air outside and at temperatures ranging from
70° to 105°C. They found the activation energy of oxidation in the induction period to be 80
KJ/mol. Using their experimental values, an induction time for medium density polyethylene of
12 years was extrapolated at a typical in-service temperature of 25°C. This value is consistent
with the approximately 20-year induction time estimated for 25-year old HDPE water and milk
bottles exhumed from a landfill. Milk and water bottles are one of a few commercial HDPE
products that do not contain antioxidants because of their limited shelf life. The exhumed bottle
materials were considered to show no signs of degradation since their yield stress, yield strain,
and modulus values had not changed significantly from those measured for new milk and water
bottles. However, there was a decrease of approximately 30% in the break strength and break
elongation values, signifying that the induction stage was essentially completed and degradation
had begun.
DRAFT - DO NOT CITE OR QUOTE 2-71
-------
The end of the induction stage signifies the onset of relatively rapid oxidation. This is the third,
and final, stage in GM degradation. Oxidation proceeds more rapidly because the free radicals
increase significantly via the decomposition of hydroperoxide. One of the free radicals is an
alkyl radical, which represents polymer chains that contain a free radical. In the early stage of
acceleration, cross-linking occurs in these alkyl radicals due to oxygen deficiency. The physical
and mechanical properties of the material subsequently respond to such molecular changes. The
most noticeable change is in the melt index, since it relates to the molecular weight of the
polymer. In this stage, a lower melt index value is detected. In contrast, the mechanical
properties do not seem to be very sensitive to cross-linking. The tensile properties (stress, strain
and modulus) generally remain unchanged or are undetectable. As time proceeds further, and
oxygen continues to be available, the reactions of alkyl radicals change to chain scission. This
causes a reduction in molecular weight. In this stage, the physical and mechanical properties of
the material change according to the extent of the chain scission. The melt index value reverses
from the previous low value to a value higher than the original starting value signifying a
decrease in molecular weight. As for tensile properties, break stress and break strain decrease.
Tensile modulus and yield stress increase and yield strain decreases, although to a lesser extent.
Eventually the GM material becomes brittle in that the tensile properties change significantly and
engineering performance is compromised, as described previously. This signifies the end of the
so-called service life of the GM.
Although arbitrary, researchers have assumed that the end of service life of a GM material occurs
when the relevant engineering properties reduce to 50% of the initial values. This is commonly
referred to as the half-lifetime, or simply the half-life. The specific property could be yield
stress, yield strain, or modulus of HDPE or the comparable break properties of resins that do not
show a pronounced yield point. It should be noted that even at its half-life the GM still exists
and can function albeit at a decreased performance level. Using the previously mentioned
Viebke et al. (1994) aging data, the half-life of unstabilized polyethylene has been estimated to
be approximately 440 years at an in-service temperature of 25°C (Hsuan and Koerner, 2002).
Considering the three stages of GM oxidation, the anticipated service life for commercially-
available HDPE GMs will be measured in terms of at least several hundred years. Other types of
GMs, particularly those with greater amorphous phase material, may have different service lives
from that for HDPE GMs. Great care should be used in specifying GM materials to require
products that, through polymer type, additive (e.g., antioxidant) packages, physical robustness,
etc., are capable of achieving as long a service life as possible.
2.5.2.10.2 GCLs
Little information currently exists on the service life of GCLs. Adequately protected and absent
of external degradation mechanisms, the service life of bentonite is indefinitely long. However,
long-term bentonite degradation is a concern if there is potential for cation exchange. In
addition, both durability and chemical compatibility are issues with respect to the reinforcing
fibers or yarns of GCLs placed on sideslopes. While the EPA test plots described by Daniel
(2002) and summarized in Section 7.4.5 go far to show the validity of such GCL reinforcement,
the performance of this reinforcement over a 30 or 100-year time frame is unknown.
DRAFT - DO NOT CITE OR QUOTE 2-72
-------
2.5.2.10.3 CCLs
For CCLs the anticipated service life is also difficult to assess, generally not from the perspective
of the soil particles themselves, but for the necessary association of the soil particles with water.
Clearly, the soil particles of a CCL will last for geologic time. However, if the CCL barrier
material should desiccate or suffer freeze-thaw cycling, its hydraulic conductivity will be
compromised. If a CCL is protected from freeze-thaw and other environmental effects, and not
subjected to excessive differential settlements, its anticipated service life is indefinitely long
(Mitchell and Jaber, 1990).
The lifetime of a CCL is clearly material and site specific. Factors that can impact the service
life of CCLs are summarized in Table 2-8.
Table 2-8. Factors affecting the anticipated service life of CCLs.
Factors Promoting a Longer CCL Service Life
Use of clayey sand or soil-bentonite mixture
Placement and compaction of soil at a relatively
low water content (e.g., on line of optimums)
Placement of CCL beneath 1 to 2 m or more of
cover soil
Protection against desiccation provided by a
GM or other type of vapor barrier
Climate with high rainfall year-round and
light to moderate drought periods of short duration
Cool climate that minimizes ET
Factors Leading to a Shorter CCL Service Life
Use of highly plastic clay
Placement and compaction of soil at a relatively
high water content (e.g., much wetter than line of
optimums)
Placement of CCL beneath less than 1 m of cover
soil
No GM or other vapor barrier provided
Climate with highly variable rainfall and with
prolonged droughts occasionally occurring
Climate with periods of year with warm temperature
and high ET or periods with freezing temperatures
2.5.3 Composite Hydraulic Barriers
A cover system with a GM/GCL, GM/CCL, or GM/GCL/CCL composite barrier allows
significantly less percolation compared to the same cover system with a GM, GCL, or CCL
barrier alone (see Section 2.5.2.3). The GM component provides protection to the underlying
GCL or CCL. The GM prevents penetration of plant roots and burrowing animals into the GCL
or CCL in most applications. The GM also protects the GCL or CCL from desiccation. The
GCL or CCL, in turn, serves to reduce the rate of leakage through occasional imperfections in
the GM.
DRAFT - DO NOT CITE OR QUOTE
2-73
-------
2.5.3.1 Prompt Placement of Overlying Materials
An interesting aspect of construction of a GM/CCL composite is that the work is generally
performed by two separate contracting organizations. The CCL is usually constructed by an
earthwork contractor and the GM is often installed by a geosynthetics installer. They rarely are
the same organizations. Thus, timing and coordination can be a challenge. To protect the CCL
from desiccation, freezing, and other stressors, the GM should be placed over the CCL as soon as
possible after the final lift of CCL is placed and accepted. In turn, after the GM in installed,
overlying layers (soil and geosynthetics) should be placed as quickly as reasonably possible.
However, all too often, days, weeks or even months pass after completion of the CCL and before
GM placement, and a similar time lag can occur with respect to the placement of overlying
materials. During this gap in construction activity, the CCL must be protected. This is difficult
since the CCL can desiccate even if left exposed for only a few days. For short-term protection,
the completed CCL should be covered by a 0.15 to 0.3 m or even thicker layer of clayey soil that
is periodically moistened and then stripped away just prior to placement of the GM.
With a GM/GCL composite liner, the GCL also should be covered with a GM as soon as
possible after installation. For GCLs, the biggest concern is that of pre-mature hydration.
A particular problem with GM/CCL composite liners is desiccation of the CCL when the GM
has been placed and left exposed (not covered with soil). Data reported by Bowders et al. (1997)
show that the exposed GM component can heat and cause desiccation of underlying clay soils
over a period of a few weeks. Desiccation occurred more rapidly with black-surfaced GMs than
with white-surfaced GMs since white-GMs reflect radiant heat, which decreases their surface
temperature. To minimize the potential for CCL desiccation, it is recommended that the GM be
covered as quickly as reasonably possible, which typically will mean that it not be left exposed
for more than several days to a few weeks prior to covering with soils. Consideration should
also be given to using light colored GMs.
If a GM/GCL composite barrier is used, the GM should also be covered as quickly as reasonably
possible, not so much over concern related to desiccation of the GCL, but, rather, over concern
related to the need to apply overburden pressure to the GCL to prevent bentonite extrusion.
2.5.3.2 Intimate Contact
Regarding intimate contact of a GM with an underlying CCL, the surface of the CCL should be
smooth rolled with a steel-drummed roller before the GM is placed, and the incidence of
wrinkles, or waves, in the GM should be minimized. Wrinkles form in the GM after initial
placement and subsequent heating during the day. At night, as the temperature declines, the GM
contracts, and the wrinkles are reduced (provided too much slack is not installed in the seamed
system). Wrinkles are more pronounced in the stiffer and thicker GMs (e.g., HOPE), but
wrinkles occur in all types of GMs because their expansion/contraction characteristics are largely
the same (Koerner, 1998). The issue with wrinkles is not that they form when the GM heats and
expands, but, rather, that as cover soils are placed on the GM the wrinkles may be trapped,
reducing contact between the GM and the underlying material. The trapped wrinkles may also
fold over, inducing stresses in the GM.
DRAFT - DO NOT CITE OR QUOTE 2-74
-------
To limit the trapping of wrinkles, cover soils should not be placed over GMs when excessive
wrinkles are present. Thus, cover soil placement should occur from daybreak until a time when
daytime heating causes wrinkles to develop. Cover soil placement can also be performed at
night. If night placement occurs, however, special precautions are needed to assure worker
safety, and intensified CQA monitoring should be conducted in recognition of the low light
conditions.
To reduce wrinkle formation, white-surfaced GMs may be considered. White-surfaced GMs
reflect more radiant heat than black-surfaced GMs, and, thus maintain a lower temperature than
black-surfaced GMs. Consequently, white-surfaced GMs experience less thermal expansion,
such that wrinkle heights are reduced by approximately one-half (Koerner and Koerner, 1995).
Since sunlight exposure is less of a factor with white-surfaced GMs, backfilling can continue
longer into the day for this GM type than for black-surfaced GMs.
On long sideslopes, it may be preferable to use textured GM rather than smooth GM to decrease
the size of GM wrinkles that develop, especially near the slope toe. Giroud (1994) has shown
analytically that GM wrinkles are shorter and spaced closer together when the shear strength
between the GM and the underlying material is increased. Therefore, based on analysis, the use
of textured, rather than smooth, GM decreases the potential for large wrinkles to form.
For GM/GCL and GM/GCL/CCL composite barriers, lateral transmission of liquid in the upper
GT of the GCL has been evaluated by Harpur et al. (1994) and found to be of little concern.
Apparently, as the bentonite hydrates it fills in, or extrudes through, the voids of the GT, greatly
decreasing the transmissivity of the GT adjacent to the GM. This, however, gives concern in
another respect. That is the possibility of decreasing the shear strength of the GM/GCL
interface. Proper direct shear testing and slope stability analyses are required when this type of
composite barrier is on steep sideslopes.
2.5.4 Construction
The manufacture, installation, QC, and CQA of GMs and GCLs and the construction, QC, and
CQA of CCLs are discussed in detail by Daniel and Koerner (1993, 1995). That detailed
discussion is not repeated herein.
In brief, GM and GCL hydraulic barriers are manufactured in panels of certain widths and
lengths. GM panels are connected by seaming using thermal processes (extrusion or fusion
seaming) for HDPE, VFPE, PVC, fPP, or fPP-R GMs or chemical processes (chemical fusion or
adhesive seaming) for fPP, fPP-R, and PVC GMs.
GCL panels are connected by overlapping. Often, dry powdered or granular bentonite is placed
within the overlap, and this practice is recommended. For GM-supported GCLs, the GM is
welded in the field. Most specifications for GCL installation require that the GCL be covered
before it becomes hydrated, and this practice is also recommended. It is common practice not to
deploy more GCL than can be covered before a rainstorm could develop.
CCLs are constructed by processing a soil and then compacting it with a certain applied energy
to a specified range of moisture contents and dry densities. The selection of moisture contents
DRAFT - DO NOT CITE OR QUOTE 2-75
-------
and dry densities for construction specifications should not be done arbitrarily but, rather, should
be based on the results of laboratory hydraulic conductivity tests performed on samples of the
proposed soil material. The resulting compaction criteria may then be narrowed based on other
engineering considerations, such as shear strength and shrinkage potential. The recommended
procedure is described by Daniel and Benson (1990), and Daniel and Koerner (1993,1995) and
has more recently been updated by Benson et al. (1999). The approach described by Daniel and
Wu (1993) is recommended for establishing appropriate moisture content-density criteria that
will ensure both low as-built hydraulic conductivity and good resistance to desiccation cracking.
Heavy, footed compactors with large feet that fully penetrate a loose lift of soil are ideal. Rollers
with feet that fully penetrate a loose lift of soil pack the base of a new lift into the surface of the
previously-compacted lift, which helps to bond lifts together. The long feet also help to break
down and remold clods of soil over the full thickness of a lift. Recommended compactor
specifications include a minimum mass of 18,000 kg and minimum foot length of 180 to 230 mm
(but the foot should have a length no smaller than the thickness of a loose lift). However, in
many landfill cover systems it is simply not possible to use such heavy compactors because the
foundation (underlain by waste at shallow depth) may not be adequate to support the weight of
the equipment. Lighter-than-ideal equipment will need to be used in such cases. To compensate
for the light weight, it may be necessary to use thinner lifts and more passes of the compactor.
When a gas collection layer is overlain by a CCL, the first lift of the CCL is sometimes
compacted with a somewhat thicker lift thickness so that the feet of the compactor don't
penetrate though the CCL and damage the underlying materials. Alternatively, the first lift of the
CCL is sometimes compacted to its specified maximum thickness with compactors having
shorter feet, rubber-tired equipment, or other equipment. This first lift is generally required to
meet compaction criteria, but may not be required to meet a permeability criterion (i.e.,
laboratory or field permeability testing of the first lift of CCL may not be required).
Soil-bentonite liners can often be compacted with rubber-tired or smooth-drum rollers. Soil-
bentonite mixtures do not develop clods, and densification of the soil is often the primary
objective with soil-bentonite liners. However, rollers with fully-penetrating feet may be
effective in bonding soil-bentonite lifts.
After compaction of a lift, the soil should be protected from desiccation and freezing.
Desiccation can be minimized in several ways: the lift can be temporarily covered with a sheet of
plastic (but one should be careful that the plastic does not heat excessively which can lead to
drying of the clay), the surface can be smooth-rolled to form a relatively impermeable layer at
the surface, or the soil can be periodically moistened. For temporary protection against freezing,
the CCL lift can be covered with a layer of clayey soil. Protection of a completed CCL was
discussed in Section 2.5.3.2.
2.5.5 Maintenance
Maintenance is discussed in Chapter 9. Since the hydraulic barrier is overlain by the surface,
protection, and drainage layers, hydraulic barrier maintenance is generally not needed unless the
cover soils and drainage layer are breached due to erosion or there are problems with slope
instability.
DRAFT - DO NOT CITE OR QUOTE 2-76
-------
2.5.6 Monitoring
Monitoring is discussed in Chapter 8. If the cover system water balance is being assessed, the
moisture content or matric potential at the top and bottom of the hydraulic barrier may be
monitored. Percolation through the hydraulic barrier may also be monitored.
2.6 Gas Collection Layer
A gas collection layer may be necessary beneath a cover system hydraulic barrier if the
underlying wastes generate gases or emit volatile constituents. The primary function of the gas
collection layer is to convey gas to some outlet (e.g., passive gas vents, active gas wells).
Collection of gases beneath a barrier can enhance cover system slope stability (see Section
6.2.2.2 and 7.7) and reduce the potential for gas emissions and lateral migration.
2.6.1 General Issues
For wastes that generate gases or emit volatiles, some type of gas management system is
required. Passive systems that rely on periodic gas vents typically require a gas collection layer
to prevent the buildup of gas pressures in the waste and beneath the hydraulic barrier.
Depending on gas generation rates, extraction well spacing, the presence or absence of horizontal
gas trenches, the air permeability of the waste, and other factors, a gas collection layer may or
may not be needed when using active gas extraction systems. However, a continuous gas
collection layer tapped periodically by relatively shallow vent pipes is the recommended
approach for many situations.
For MSW landfills, which may generate significant quantities of gas, control of gas beneath
cover systems with a GM, GCL, or composite barrier is especially important. If gas is not
properly managed, the gas may migrate through the subsurface (as opposed to venting to the
atmosphere), causing potential safety hazards in enclosed areas, on adjacent properties, etc.
Subsurface gas migration may also lead to adverse groundwater quality impacts due to diffusion
of volatile constituents from the gas phase to groundwater. Moreover, uncontrolled gas buildup
beneath a GM, GCL, or composite barrier will produce uplift pressure that will either cause GM
bubbles (or "whales") to occur, displacing the cover soil and appearing at the surface (Figure 7-
23), or cause a decrease in the normal stress between the GM or GCL and the underlying
material. The whales can cause excessive deformations in the cover system components. The
authors are aware of at several cases where an HDPE GM was deformed past its yield strain
when a whale developed. At several facilities, the latter effect (i.e., decrease in normal stress)
led to slippage of the GM and overlying cover materials creating high tensile stresses evidenced
by compression ridges in the cover soil and folding of the GM at the slope toe and tension cracks
in the cover soil near the slope crest. One example of a cover system stability problem caused by
gas pressures is described in Section 7.7. Briefly, gas generated in a MSW landfill uplifted the
GM barrier of a cover system and resulted in the GM and overlying materials moving downslope
over a GT. Though the landfill had vertical gas extraction wells, the upper portion of the wells
was not perforated. As a consequence, gas accumulated beneath the cover system, generating
uplift pressures on the underside of the GM.
Gas collection layers should be designed to provide free-flow of gas to outlets. Methods for
calculating the maximum flow rate in a gas collection layer are presented in Section 5.3. The
DRAFT - DO NOT CITE OR QUOTE 2-77
-------
allowable flow rate of a gas collection layer can be calculated as described in Section 2.6.2.3.
Outlet design is discussed in Section 2.6.2.4.
The need for a soil or GT filter between the gas collection layer and overlying hydraulic barrier
should be evaluated. For example, a GT is often used between a CCL and a granular or GN gas
collection layer to prevent CCL material from being pushed into the gas collection layer during
construction and retain the CCL particles should percolation occur. In this application, the GT is
serving as a separator and a filter. A GT filter may also be required between a GCL and a gas
collection layer to prevent downward extrusion of hydrated bentonite. The design of soil and GT
filters is presented in Section 4.7.
2.6.2 Elements of Design
Important questions that typically need to be addressed when considering the design of the gas
collection layer include:
• What materials are available to construct the gas collection layer?
• What thickness of gas collection layer material is needed?
• What is the maximum design flow rate and the allowable flow rate in the drainage layer?
• How should gas collection layer transitions and outlets be designed?
• How should the gas collection layer be constructed?
• What type and frequency of maintenance should be employed?
• What type and frequency of monitoring should be employed?
2.6.2.7 Materials
Like drainage layers (see Section 2.4.2.1), gas collection layers may be constructed of granular
materials or geosynthetics. The material used should have adequate gas conductivity to
minimize the build up of gas pressures beneath the barrier and adequate gas transmissivity to
convey the design gas flow rate.
2.6.2.1.1 Granular Materials
Granular gas collection materials are normally composed of relatively clean sand or gravel.
When a granular material is used, a separation or protection layer (typically a GT) may be
needed between the granular material and the overlying barrier.
Specifications for granular materials often require:
• no more than 5% (dry-weight basis) of material passing the No. 200 sieve;
• a maximum particle size on the order of 25 to 50 mm;
• a GT cushion may be required between the GM and granular material to protect the GM
from damage (e.g., deep scratches, puncture);
• restrictions on gradation, stated in terms of allowable percentages for specified sieve
sizes (these restrictions may exist for various purposes);
DRAFT - DO NOT CITE OR QUOTE 2-78
-------
• restrictions on the angularity of the material, if the material will interface with
geosynthetics, which are vulnerable to puncture by large, sharp objects (or, alternatively,
a GT cushion may be employed);
• that no deleterious material be present; and
• a minimum hydraulic or gas conductivity.
Gas conductivity of granular material is occasionally measured directly in the laboratory using
techniques such as those described by Scanlon et al. (1999). However, more often it is estimated
from the soil hydraulic conductivity as:
(Eq. 2.24)
where: kg = gas conductivity (m/s); k = hydraulic conductivity (m/s); pg = gas density (kg/m3); pw
= water density (kg/m3); (j,g = gas viscosity (kg/m/s); and u.w = water viscosity (kg/m/s).
Laboratory hydraulic conductivity testing of granular materials is discussed in Section 2.4.2.1.1.
Gas conductivities are typically 20 times less than hydraulic conductivities because gas density is
approximately three orders of magnitude less than water density and gas viscosity is
approximately 50 times less than water viscosity. Because the gas permeability of a material
decreases as its pore space becomes filled with water, gas collection layers should be designed to
remain relatively dry and should be installed in a relatively dry state.
2.6.2.1.2 Geosynthetics
A range of geosynthetics, such as those described in Section 2.4.2.1.2, can be used for the gas
collection layer. Like granular gas collection layers, a geosynthetic gas collection layer should
meet filter criteria with the overlying hydraulic barrier. Furthermore, if a GM hydraulic barrier
overlies a GN or core gas collection layer, a GT may be required between the collection layer
and GM to provide higher interface friction on steep sideslopes and, possibly, reduce
deformation-related intrusion of the GM into the collection layer and/or protect the GM from
puncture or other damage by the collection layer.
Specifications for geosynthetic gas collection layers often require:
• resin and additive requirements;
• minimum thickness;
• minimum mass per unit area;
• specified density;
• minimum air transmissivity at a specified normal stress and gradient;
• minimum strength requirements to survive installation;
• if the gas collection material is a GN or core, inclusion of a GT above the material, if
necessary, to increase interface friction, reduce deformation-related intrusion of an
DRAFT - DO NOT CITE OR QUOTE 2-79
-------
overlying hydraulic barrier into the material and/or protect the hydraulic barrier from
puncture or other damage by the drain; and
• if the gas collection material is a GN or core, inclusion of a GT filter below the material.
Gas transmissivity of geosynthetics is occasionally measured directly in the laboratory (e.g.,
Koerner (1997) presents data for needlepunched nonwoven GTs), but is more often estimated
from the geosynthetic hydraulic transmissivity using Eq. 2.23 with the gas transmissivity, 9S
1 1
(m /s/m), substituted for kg and the hydraulic transmissivity, 9h (m /s/m), substituted for k.
Because the gas transmissivity of a material decreases as its pore space becomes filled with
water, gas collection layers should be designed to remain relatively dry and should be installed in
a relatively dry state.
2.6.2.2 Thickness of Granular Layers
The recommended minimum thickness of a granular gas collection layer is usually 0.3 m. This
allows sufficient thickness for ease of construction. With extremely careful control of thickness,
it is possible to construct even thinner granular gas collection layers (down to a thickness of
about 0.15 m), but granular gas collection layers thinner than 0.3 m are not very common.
2.6.2.3 Required Flow Capacity
Similar to a drainage layer, a gas collection layer, either granular material or geosynthetic, can be
designed using Eq. 2.21. Methods for calculating the maximum flow rate are presented in
Section 5.3. FS values should be selected considering the uncertainties in the various design
variables and the consequences of failure.
For all types of gas collection layer materials, the required hydraulic properties are evaluated
considering the material properties measured in the laboratory and reduction factors that consider
the potential for long-term clogging, deformation, etc. Eqs. 2.22 and 2.23 for drainage layer
materials can be used with Eq. 2.24 for this purpose.
2.6.3 Gas Collection Layer Outlets
As previously discussed, gas or vapors collected in the gas collection layer should be conveyed
to an outlet, which is typically a vertical riser pipe or vent. Since each outlet requires penetration
of the hydraulic barrier, the number of outlets should be limited. Ideally, outlets should be
located at high points within the cover system, although this is not always possible. Connections
between gas outlets and the hydraulic barrier should be carefully designed to prevent water
infiltration through and around the gas outlets and to accommodate differential settlements
between the outlets and the barrier. The authors are aware of connections that were damaged
due to differential settlement. For example, as described in Section 7.5, cover system GM boots
around the gas well penetrations at a MSW landfill were not designed to accommodate
settlement of the waste, which would cause downward displacement of the GM barrier relative to
the wells. Within about one year after construction, 0.3 to 0.9 of differential settlement had
occurred and the GM boots had torn from the GM barrier. The problem was resolved by
replacing the gas extraction well boots with new expandable boots that could elongate up to 0.3
m and could also be periodically moved down the well.
DRAFT - DO NOT CITE OR QUOTE 2-80
-------
2.6.4 Construction
The construction, QC, and CQA of granular gas collection layers and the manufacture,
installation, QC, and CQA of geosynthetic gas collection layers are discussed in detail by Daniel
and Koerner( 1993, 1995).
In brief, granular material is usually loosely dumped from a truck and spread with a low-ground
pressure bulldozer. Low-ground pressure equipment is used to minimize the generation of fines.
Granular gas collection layers are generally not compacted.
Geosynthetic drainage layers are manufactured in panels of certain widths and lengths. The
panels are placed in the field and connected by overlapping, seaming, tying, interlocking, or
other means.
When a gas collection layer is overlain by a CCL, the first lift of the CCL is sometimes
compacted with a thicker lift thickness so that the feet of the compactor don't penetrate though
the CCL and damage the underlying materials. Alternatively, the first lift of the CCL is
sometimes compacted to its specified maximum thickness with compactors having shorter feet,
rubber-tired equipment, or other equipment. This first lift is generally required to meet
compaction criteria, but may not be required to meet a permeability criterion (i.e., laboratory or
field permeability testing of the first lift of CCL may not be required).
2.6.5 Maintenance
Maintenance is discussed in Chapter 9. Since the gas collection layer is overlain by the surface,
protection, and drainage layers and the hydraulic barrier, gas collection layer maintenance is
generally not needed unless there are problems with slope instability.
2.6.6 Monitoring
Monitoring is discussed in Chapter 8. Depending on the design of the gas collection system, the
flow rates and chemistry of gas removed from the gas collection layer may be monitored.
2.7 Foundation Layer
The foundation layer is the lowermost component of the cover system. The primary functions of
the foundation layer are to provide grade control for cover system construction, adequate bearing
capacity for overlying layers, a firm subgrade for compaction of overlying layers, and a smooth
surface for installation of any overlying geosynthetics. In some applications, the foundation
layer may be designed to attenuate the potential effects of waste differential settlements on the
cover system components (e.g., the foundation layer may be required to have a certain
thickness). If the foundation layer material is granular, the layer may also serve as a gas
collection layer.
2.7.1 General Issues
Waste receives its final mechanical compactive effort during placement of the foundation layer.
To minimize post-construction settlement, and especially differential settlement, of the cover
DRAFT - DO NOT CITE OR QUOTE 2-81
-------
system, the foundation layer should be heavily proofrolled with large compactors. However,
even a large compactor will not compact waste below a depth of about 1 to 2 m.
To compact the waste to greater depths, as may be required when warehouses or other structures
are constructed on a cover system, the foundation subgrade may be proofrolled before the
foundation layer is placed or preload fill or deep dynamic compaction may be used. A detailed
description of the dynamic compaction method is presented by Mayne et al. (1984). With deep
dynamic compaction, a large weight (usually a concrete block) is dropped from a height of many
meters transmitting high energy to the ground surface. The impact of the weight compacts the
underlying materials and collapses voids, causing deformation in both vertical and horizontal
directions. Dynamic compaction is carried out in several passes, with the weight dropped in a
predetermined grid pattern during each pass. The resulting craters are eventually filled with soil
and the surface is proofrolled.
The depth of influence of the technique depends on the physical and dynamic properties of the
material to be compacted, the location of the groundwater table, and other factors. As a general
rule, the depth of influence for soils (not necessarily solid waste) can be estimated from the
following empirical equation:
Di = a (W H)°5 (Eq. 2.25)
where: a = empirical constant between 0.3 to 1 (m/tonne)0'5, with the specific value depending on
soil grain size distribution and degree of saturation; D; = depth of influence (m); W = mass of the
falling weight (tonne); and H = height of the falling weight (m). It has been estimated that for
soil densification, the densification is substantial down to a depth equal to about D/2 (Mayne et
al., 1984), beyond which it decreases.
2.7.2 Elements of Design
Important questions that typically need to be addressed when considering the design of the
foundation layer include:
• What materials are available to construct the foundation layer?
• What thickness of foundation layer material is needed?
• How should the foundation layer be constructed?
• What type and frequency of maintenance should be employed?
• What type and frequency of monitoring should be employed?
2.7.2.1 Materials
Materials most often used for the foundation layer include on-site or locally available soils. For
landfills, daily or intermediate cover soil already in place is sometimes used for all or a portion
of the foundation layer. In a few situations, waste material can be used to construct the
foundation layer. If constructed of granular material, the foundation layer may also serve as a
gas collection layer.
DRAFT - DO NOT CITE OR QUOTE 2-82
-------
2.7.2.2 Thickness
The thickness of the foundation layer is selected based on site-specific criteria. The minimum
thickness of a foundation layer is usually 0.3 m. When the foundation layer is designed to
attenuate the waste differential settlements, it may be several meters to more thick.
2.7.3 Construction
The foundation layer may be placed and compacted using procedures for structural fill or may
have no specific compaction criteria. At a minimum, the foundation layer is generally heavily
proofrolled with large compactors, as described in Section 2.7.1. As many load repetitions as
practical may be used so that stresses are felt as deeply as possible in the waste mass.
2.7.4 Maintenance
Maintenance is discussed in Chapter 9. Since the foundation layer is overlain by the other cover
system components, foundation layer maintenance is generally not needed unless there are
problems with slope instability.
2.7.5 Monitoring
Monitoring is discussed in Chapter 8. If the cover system water balance is being assessed, the
foundation layer moisture content or matric potential may be monitored. Percolation through the
foundation layer may also be monitored.
2.8 Examples of Cover Systems for Different Applications
Cover systems can be constructed with a wide variety of configurations of soil and geosynthetic
layers to satisfy project-specific design criteria. A few examples used on specific projects are
presented below. Additional examples of cover system cross sections can be found in Koerner
and Daniel (1997).
Figure 2-17 illustrates two different hydraulic-barrier type of covers systems for a MSW landfill,
one with a CCL hydraulic barrier and the other with a GM/CCL composite hydraulic barrier.
For either example, a GCL can be considered as an alternate to the CCL. The choice of the
underlying soil material, CCL or GCL, is controlled primarily by the how these materials
respond to the anticipated differential settlements, wet-dry cycles, freeze-thaw cycles, and shear
stresses and economics. The mechanical and hydraulic properties of CCLs and GCLs were
discussed previously in Section 2.5. Soil thicknesses for this type of cover system will vary
based on project-specific conditions.
Figure 2-18 presents an ET-barrier type of cover system for a MSW landfill in an arid setting.
Design of the ET-barrier type of cover system is discussed in Section 3.2. Cover systems
constructed at arid sites often require surface layers that are more resistant to erosion than
vegetated topsoil. As discussed in Section 2.2.2.2, gravel-soil mixtures, gravel veneers, riprap,
and other materials may be used as surface layer material for this purpose. MSW landfills
constructed in arid environments may need a gas collection layer beneath the ET barrier
depending on the gas generation rates in the landfill and the efficiency of any gas collection
system. Soil thicknesses will vary based on project-specific conditions.
DRAFT - DO NOT CITE OR QUOTE 2-83
-------
Figure 2-19 presents the cover system for a low-level radioactive waste landfill with a minimum
design life of 200 years. The cover system for a low-level radioactive waste disposal facility is
typically designed with a higher level of protection than cover systems for MSW and hazardous
waste landfills. For the cover system in Figure 2-19, the protection layer includes a thick
biointrusion layer to minimize the potential for exposure of animals and plants to waste. It also
incorporates a GM/GCL/CCL composite hydraulic barrier. As for cover systems over MSW and
FEW landfills, soil thicknesses will vary based on project-specific conditions.
Figure 2-20 shows the lightweight cover system used as part of the remediation of an
uncontrolled dumpsite containing HW. The site is in a marsh. The low bearing capacity of the
foundation soil and waste at the site necessitate the use of this type of cover system. As
described in Section 6.6, if the waste to be covered is a quasi-liquid (e.g., a sludge), the design of
the cover system is often different. In such cases, the waste strength is increased (by physical
solidification, dewatering, or other means), the cover system is reinforced, and/or a lightweight
cover soil that includes a GM or a GCL is used. Geotechnical design consideration for cover
systems on soft waste materials are discussed in more detail in Section 6.6.
Figure 2-21 illustrates "floating covers" for liquid or sludge waste impoundments. While GM
floating covers placed over impoundments are rarely considered "cover systems", they often
remain in place for many years and, in effect, may be designed to function as cover systems. For
this reason, liquid waste impoundment covers are mentioned here. Liquid wastes may be
covered with a GM to reduce emissions of volatile waste constituents, meet personnel safely
requirements, and satisfy aesthetic requirements. The dimensions of the GM are proportioned
when the impoundment is empty, if there is any possibility that draining of the impoundment
may occur. To keep the central portion of the cover quasi-stable, expanded polystyrene (EPS)
floats may be attached to the underside of the GM in a pattern that creates a stiffened central
portion (Gerber, 1984). The slack is accumulated on the sides of the impoundment where it is
accommodated by an arrangement of parallel floats with a sand tube welded to the upper side of
the GM (Figure 2-21 (a)). When the trough that is created by the floats and sand tube fills with
rainwater, the water can be pumped from the GM surface. An alternative to this type of slack
accommodating system is the tensioned-membrane approach illustrated in Figure 2-2l(b). Here
the GM is configured with tensioned lines such that weights in adjacent steel stanchion posts
move up or down as the liquid level falls or rises. For the cases illustrated in Figure 2-21, wind
loads can induce significant stresses, and GM edge and connection stresses are very high.
Because of this, Koerner (1998) recommends that GM covers meet the minimum strength values
given in Table 2-7 for a very high degree of installation survivability. Furthermore, since the
GMs are continuously exposed to the environment, they require excellent resistance to ultraviolet
degradation. Favored in view of these two requirements are fPP-R, CSPE-R, and ethylene
interpolymer alloy-reinforced (EIA-R) GMs.
DRAFT - DO NOT CITE OR QUOTE 2-84
-------
I
0,45 m
0.30m
4
0.45m
1
Soil
Granular Soil or Geosynthetic
CCL
< 10-8m/s)
; ' Gas Collection/Foundation Layer
Surface/
Protection Layer
Drainage Layer
Hydraulic
Barrier
Waste
(a)
I
0.6m
1
4
0.45 m
1
Soil
="\J~tT\y\7"\ J\3^S\J\7\j\J\f\J\J\j\J^JTJ~\^S\J^7\S*£r'
CCL
(k<1Q-7m/s)
•"-.• " Gas Collection/Foundation Layer . ; •' "
- ~" *""-v"-u'i--»"-t'""k"-ir~' "*
^ ^~" " .*" ™ ^>J -
I
! Surface/
Protection Layer
r Drainage Layer
Hydraulic
Barrier
»
Waste
(b)
Figure 2-17. Examples of Hydraulic Barrier-Type of Cover Systems for MSW Landfills:
(a) Cover System with CCL Hydraulic Barrier; (b) Cover System with
GM/GCL Composite Hydraulic Barrier.
DRAFT - DO NOT CITE OR QUOTE
2-85
-------
0.15 rn
0.9m
Soil
Surface/
Protection Layer
ET Barrier
s nollRction/Foundation Layer :.- ;;.: ] V_GT separator
^^^^^*r^ ^f- 2T:.*?t*
T^r *--.>?*•-
Waste
Figure 2-18. Example of ET Barrier-Type of Cover System for MSW Landfills.
2.95m
0.15m
0.55m
0.15m
0.9m
0.3m
Topsoil
Vegetative Soil Layer
Granular Filter
Surface/Protection
Layer
Biointrusion Barrier
Granular Soil
0.6m
0.3m
CCL
(k<10-9m/s)
/
Foundation Layer
Drainage
Layer
GT Cushion
GM
M3CL
Hydraulic Barrier
Waste "r--_
Figure 2-19. Example of Cover System for a Low-Level Radioactive Waste Landfill.
DRAFT - DO NOT CITE OR QUOTE
2-86
-------
0.3 m
0.3 m
General Fill
Regraded Existing Fill
Surface/Protection Layer
GC Drainage Layer
Hydraulic Barrier (GM and GCL)
GC Gas Collection Layer
Geosynthetic Reinforcement
(where needed)
Foundation Layer
Figure 2-20. Example of Lightweight Cover System for a HW Remediation Site.
DRAFT - DO NOT CITE OR QUOTE
2-87
-------
(a)
Tensioned Line
Weight
GM Floating on Liquid
Liquid
EPS Float
GM Floating on Liquid
Sand Tube
Concrete
Anchorage
Liquid
(b)
GM Liner Component
Figure 2-21. Examples of Floating "Cover System" for HW Impoundments: (a) GM with
Tensioned Lines; and (b) GM with Floats and Sand Tubes.
DRAFT - DO NOT CITE OR QUOTE
2-88
-------
Chapter 3
Alternative Design Concepts and Materials
3.1 Introduction
As previously mentioned in Section 1.3, RCRA and CERCLA regulatory requirements provide
flexibility for innovation and alternatives in cover system design. The regulatory mechanism for
approval of an alternative design or material typically includes a demonstration of technical
equivalence. The alternative must perform in a manner that is equivalent or superior to the
design or material it replaces. Depending on the function of the proposed cover system
alternative, the demonstration of technical equivalence may include an evaluation of water
percolation through the cover system, gas emission rate, erosion potential, and/or long-term
performance (e.g., ability to accommodate foundation settlements, service life). Some of the
alternative design concepts and materials discussed in this chapter have met this equivalency
criterion on a project-specific basis and have been employed in cover systems for a limited
number of landfills and contamination source areas.
The two alternative cover system design concepts discussed in this chapter (with a performance
goal of preventing precipitation from percolating through the cover system) are based on either:
(i) the evapotranspiration (ET) barrier principle; or (ii) the capillary barrier principle. Cover
systems with an ET or capillary barrier are generally best suited for semi-arid and arid climates
with minimal snowpack, and capitalize on the naturally occurring low precipitation rates and
high potential evapotranspiration (PET) rates in these climates. Arid sites generally receive less
than 250 mm of annual rainfall with evaporation exceeding rainfall and sparse vegetation, and
semi-arid sites have a mean annual precipitation between 250 and 500 mm and are typically
vegetated with grasses (Lincoln et al., 1982). The extent of arid and semi-arid lands in the U.S.
is shown in Figure 3-1. In wetter climates, these alternative cover system design concepts are
generally not as effective as designs with hydraulic barriers since the fine-grained soil layers
used to store infiltrating water in the alternative designs would have to be relatively thick to
provide adequate water storage capacity, and water migrating into the lower regions of these soil
layers may not be easily removed by ET. The alternative design concepts differ from designs
with hydraulic barriers alone in that they are intended to emphasize the following:
• unsaturated hydraulic conductivities of the soil components;
• low hydraulic conductivity of fine-grained soil layer(s), even at high degrees of soil
saturation;
• relatively high water storage capacity of fine-grained soil layer(s) with eventual removal
of stored water primarily by ET;
• increased transpiration through the use of diverse native vegetative; and
• ease of construction and/or substantial cost savings through the use of locally-available
materials.
DRAFT - DO NOT CITE OR QUOTE
3-1
-------
Figure 3-1. Semi-Arid and Arid Areas in the U.S. (modified from Meigs, 1953).
Because the soil layers in the alternative designs are relatively dry, they often have moderate to
high gas permeabilities and, therefore, may not provide an effective barrier to gases, if any,
generated within the landfill or contamination source area. It is important that the potential for
gas generation and the need to collect and manage gases be considered when developing an
alternative cover system design. If gas generation may occur, the collection, transmission, and,
potentially, treatment of these gases should be considered. If the facility is a MSW landfill
subject to EPA's gas collection and treatment regulations or if gas emissions through the cover
system are a concern, the facility should incorporate appropriate gas containment components.
The effect of seasonal freezing of near surface soils on lateral and downward gas migration also
needs to be addressed.
In some areas in the southwest, regulatory agencies are promoting the use of alternative cover
system designs to EPA performance criteria and guidance for MSW landfills. There is a concern
that the CCL component of a GM/CCL composite barrier in a cover system may desiccate and
crack over time, especially in semi-arid and arid climates (EPA, 1989; EPA, 1991; Suter et al.,
1993), providing little value to the cover system. As an example, in southern California,
regulators are currently allowing use of cover systems with ET barriers to close MSW landfills
constructed without a Subtitle D liner system. The cross section of an ET barrier cover system
constructed at such a landfill is shown in Figure 3-2.
The design of ET and capillary barriers is discussed in more detail below. Additional design and
construction considerations for these cover systems are presented in "Technical and Regulatory
Guidance for Design, Installation, and Monitoring of Alternative Final Covers" (ITRC, 2003),
and in "Evapotranspiration Landfill Cover Systems Fact Sheet" (EPA, 2003). These designs
DRAFT - DO NOT CITE OR QUOTE
3-2
-------
should be carefully reviewed by a person knowledgeable and experienced in unsaturated soil
moisture modeling and the design of such cover systems. Because there are uncertainties in the
design assumptions and methods and field performance data for alternative cover system designs
are limited, EPA is presenting a conservative design approach herein. Furthermore, EPA
recommends that field monitoring of these cover systems be conducted to verify that the design
assumptions and methods are appropriate. With these data, design procedures may be refined for
a given geographic area. This is already occurring in southern California, where a more unified
approach to the modeling and field monitoring of ET barriers is evolving.
1.2m
Surface/Protection Layer
0.6m
Foundation Layer (Cover Soils)
Figure 3-2. Cross Section of ET Cover System Used for a MSW Landfill in Southern
California.
Chapter 3 also discusses emerging alternative materials that can be used in lieu of the various
materials traditionally used in cover systems and described in Chapter 2. The considered
alternative materials are geofoam, shredded tires, sprayed elastomers, and paper mill sludges.
3.2 ET Barrier Design
3.2.1 Overview
As discussed in Section 1.1.2 and illustrated in Figure 1-4, ET barriers consist of a thick layer of
relatively fine-grained soil. The barrier may be overlain by a topsoil layer or surface treatment
to promote vegetative growth and reduce the potential for erosion by water or wind. Soil types
DRAFT - DO NOT CITE OR QUOTE
3-3
-------
used for construction of ET barriers include fine-grained soils such as silty sands, silts, and
clayey silts. In general, the greater the percentage of fines in a soil, the greater the water storage
capacity and thus the thinner the barrier required to store a given amount of water. As discussed
in Section 2.3.2.2.3, soils with a large fraction of clay are typically not used due to the potential
for desiccation cracking of the clay. Cracks provide preferential pathways for infiltrating water
to bypass the clay matrix and thereby bypass storage. In addition, there is somewhat less
available water for plants in clays than in silty soils (Figure 2-11).
Previous research has shown that a simple ET barrier can be effective at limiting percolation and
erosion, particularly in dry environments (Nyhan et al., 1990; Hauser et al., 1994; Nyhan et al.,
1997; Dwyer, 1998; Dwyer, 2001). The thickness of the barrier is selected, based on the barrier
soil's water storage capacity (Eq. 2.5) to retain infiltrating water until it can be removed by ET.
Saturated flow in the near surface, when it does occur, is primarily downward as the hydraulic
gradient is largely due to gravitational potential differences. Water movement deeper in the soil
profile generally occurs under an unsaturated condition. Under this condition, the hydraulic
gradient is comprised of a gravitational potential component (acting downward) and a matric
potential component (which can act either upward or downward) (see Eq. 4.11). Matric
potential gradients can be many orders of magnitude greater than the gravitational potential
gradient. Water flows in response to the total potential gradient. Since the total potential
gradient is the sum of the matric potential gradient, gravitational potential gradient, and other
gradient components (e.g., solute potential gradient) which are generally less significant and are
not considered in this guidance document, both upward and downward water movement is
possible in the unsaturated soil of an ET barrier.
As previously mentioned, ET provides the mechanism to remove stored water from the ET
barrier. Evaporation of water from the soil surface decreases the soil water content and, thus,
matric potential in the upper portion of the barrier. This results in an upward matric potential
gradient and upward flow. Plant transpiration also relies upon water potential gradients (matric
and osmotic) to remove water from the ET barrier. Figure 3-3 shows a typical variation in water
potential in the soil-plant-atmosphere system. In arid climates, the total water potential
difference between soil moisture and atmospheric humidity can exceed 100 MPa (10,000 m of
water) (Hillel, 1998). The largest portion of this overall potential difference occurs between the
leaves and the atmosphere. The larger the soil-plant-atmospheric potential gradient, the more
effective is the ET barrier. For this reason, well-vegetated ET barriers can be very effective in
semi-arid and arid regions. These regions are characterized by large potential evapotranspiration
(PET) compared to precipitation.
PET is an index that essentially represents the atmospheric "demand" for water. PET can be
calculated using a form of Penman's equation (Penman, 1948). The total calculated PET for
Tucson, Arizona from January 1987 through December 1999 was 25.71 m while the actual
precipitation during this period was only 3.61 m (http://ag.arizona.edu/azmet/). This equates to a
greater than 7:1 PET to precipitation ratio (i.e., there is a much greater demand for water by the
atmosphere and plants than can be supplied to the soil by precipitation). A monthly comparison
of PET versus precipitation for 1999 is shown graphically in Figure 3-4.
DRAFT - DO NOT CITE OR QUOTE
3-4
-------
Soil Water
(-30 kPa)
Figure 3-3. Typical Soil-Plant-Atmosphere Water Potential Variation (modified from Hillel,
1998).
ouu
f—
| 250
CL 200
O
c
B 150
TO
'5.
'o
-------
3.2.2 General Issues
A number of the same general issues that were mentioned in Sections 2.2.1 and 2.3.1 for surface
and protection layers, respectively, also apply to ET barriers. Important issues are water storage
capacity and erosion potential, since excessive erosion can cause the cover to be ineffective.
3.2.3 Elements of Design
Important questions to be addressed when designing an ET barrier include the following:
• What materials should be used to construct the barrier?
• How thick should the barrier be to store the required amount of water?
• Are materials uniform and have appropriate placement methods been determined to
minimize preferential pathways for percolation?
• What surface treatments should be applied to control erosion?
• Which plants should be established to promote transpiration and stabilize the cover
surface?
• How and at what frequency should the barrier be maintained?
• What type and frequency of monitoring should be employed?
3.2.4 Design Concept
The ET barrier design concept can be summarized in the following steps:
1. Identify the critical infiltration event(s) that may result in percolation. This generally
involves identifying the design precipitation event or series of events. Khire et al. (2000)
recommend that the meteorological record for the site be reviewed to define critical time
periods where PET less precipitation is near zero or negative. This condition should
normally occur outside the growing season (Khire et al., 2000).
2. Calculate the depth of water that must be stored in the ET barrier based on the design
infiltration event(s). For simplicity, it can be assumed that the barrier must hold all of the
precipitation occurring during the critical infiltration event(s), i.e., there is no runoff or
ET (Khire et al., 2000).
3. Characterize the unsaturated hydraulic properties of the considered fine-grained barrier
soil and calculate its water storage capacity using Eq. 2.5.
4. Calculate the minimum soil thickness required for the fine-grained soil as described in
Section 3.2.5.
5. Establish the vegetation (seed mix) to be used and any surface treatment (i.e., gravel
veneer, gravel admixture, soil nutrient supplements) to be employed. Cover system
vegetation is discussed in Sections 1.6.6 and 2.2.3. Surface treatments are described in
Section 2.2.2.2.
6. Assess the need for optional layers (i.e., gas vent layer, biointrusion layer). Optional
layers are described in Chapter 2.
DRAFT - DO NOT CITE OR QUOTE
3-6
-------
7. Establish the adequacy of the design based on:
- predictive computer modeling (Section 3.4.2),
- field data to evaluate short-term performance (Section 3.4.3), and
- natural analogs to predict long-term performance (Section 3.4.4).
3.2.5 Soil Thickness
An estimate of the required thickness of the ET barrier can be made based on the required depth
of water to be stored in the soil and the water storage capacity of the soil. The design strategy
for an ET barrier is to ensure that the storage capacity is sufficient to store the "worst-case"
infiltration quantity resulting from the critical infiltration event(s), with an appropriate factor of
safety, until the infiltration can be removed via ET.
As discussed in Section 2.3.2.2.7, the depth of water, Hw, that can be stored in a soil layer is the
product of the water storage capacity, 9SC, of the soil and the layer thickness, Hs. The storage
capacity, in turn, is a function of the soil's field capacity and permanent wilting point.
Representative values of 9SC for different soil textures were presented in Table 2-6.
In dry environments, plants commonly reduce the water content of a near-surface soil to the
permanent wilting point during every growing season (Anderson et al., 1993), making the soil's
entire storage capacity available for subsequent precipitation when ET is low and plants are
dormant. Thus, one potential scenario of the required amount of infiltration that an ET barrier
has to store annually is the total precipitation input during the dormant period(s). Another
scenario might be that created by spring snowmelt or summer thunderstorms. Both of these
design scenarios should be considered.
ET-barrier type cover systems located in temperate climates have been vegetated with perennial,
fast-growing, and deep-rooted hybrid poplar trees (Licht et al., 2001). Hybrid poplar trees have
been used for phytoremediation and have been considered for cover system applications (i.e.,
phytocaps) because they exhibit relatively high water uptake rates (e.g., 810 to 1,070 mm/yr for
tree plantations) and growth rates (e.g., 1 to 3 m/yr), develop deep root systems (2 to 3 m deep),
are easily propagated, and can be planted economically. Two cover systems with ET cover
systems vegetated with hybrid poplars are being monitored under the Alternative Cover
Assessment Program (ACAP), which is discussed in Section 3.4.3.
Generally, there is a need to incorporate a factor of safety into the design of an alternative barrier
to help offset some of the uncertainties associated with weather, in-place soil properties, and
vegetation growth. Reasonable values for these parameters should be used and a factor of safety
should be applied, at a minimum, to the required amount of water to be stored. Since there are
few field performance data available for alternative cover systems, EPA believes that the
minimum thickness of an ET barrier should be the larger of 1.25 Hs (i.e., a factor of safety of
1.25 applied to the calculated cover soil layer thickness) and 0.9 m. This factor of safety and
minimum thickness not only account for uncertainities in precipitation, modeling, and material
properties, but also allow for the possibility of long-term erosion of the surface soil. This level
of conservatism may be reduced somewhat when the performance of the alternative barrier is
DRAFT - DO NOT CITE OR QUOTE
3-7
-------
modeled using an unsaturated flow code and site-specific parameters, if the cover system is
monitored (see Chapter 8), or if a GM is used beneath the ET barrier. The latter case may apply
when an ET/GM composite barrier is used in lieu of a GM/CCL composite barrier.
As an example, during 1987 to 1999 Tucson, Arizona received from about 5.1 to 236.0 mm of
precipitation annually during December and January, when plants are typically dormant. The
average precipitation during this time period was 58.2 mm. Dividing the worst-case
precipitation value of 236.0 mm by a storage capacity of 0.15 for a silty loam soil yields a
required ET barrier thickness of 1.7 m. Applying a factor of safety of 1.25 to this thickness
yields a design thickness of 2.125 m. The above calculation method is simple, but conservative,
and doesn't take into account runoff or evaporation. When the above scenario was simulated
using an unsaturated flow model with historical weather data and assuming the silty loam soil
was initially at its wilting point, the required barrier thickness to limit percolation to less than 0.5
mm/yr during the simulation period was calculated to be approximately 0.8 m. Applying a factor
of safety of 1.25 to this thickness yields a design thickness of 1.0 m.
3.3 Capillary Barrier Design
3.3.1 Overview
As discussed in Section 1.1.2 and illustrated in Figure 1-5, capillary barriers consist of one or
more layers of finer-grained soil overlying one or more layers of coarser-grained soil. Like the
ET barrier, a capillary barrier may have a topsoil layer or surface treatment to promote
vegetative growth and reduce the potential for erosion. The finer-grained soil in a capillary
barrier has similar characteristics to the fine-grained soil used to construct an ET barrier: it is
generally a silty soil, as described in Section 3.2.1. Soil types used for construction of the
coarser-grained component range from coarse sand to cobbles.
The capillary barrier design concept relies on the differences in pore size distribution between
the upper finer-grained soil and the lower coarser-grained soil to promote retention of water in
the finer-grained soil under unsaturated flow conditions, as long as the contrast in unsaturated
properties (e.g., soil-moisture characteristics and unsaturated hydraulic conductivities) of the two
soils is sufficiently large. This can be explained as follows: at a given matric potential, a
coarser-grained soil tends to have a much lower water content than a finer-grained soil. The
hydraulic conductivities of unsaturated soils decrease exponentially with decreasing water
content because flow paths through thin films of water coating the soil particles in dry soil are
extremely tortuous. Thus, dry gravel is actually much less permeable to water than moist silty
sand. If the soils remain unsaturated, the finer-grained soil tends to retain nearly all the soil
water and the underlying layer serves as a barrier due to its dryness. The matric potential in the
finer-grained soil layer typically must approach a value near zero (i.e., saturated conditions)
before any appreciable flow occurs into the coarser-grained layer (Figure 1-5).
In contrast to ET barriers, which experience primarily vertical water flow, the primary direction
of water flow (i.e., vertical or lateral) in capillary barriers depends on whether or not the
capillary barrier is sloped. The water balance for non-sloped capillary barriers is similar to that
for ET barriers. Thus, water is removed from the finer-grained soil component of a non-sloped
DRAFT - DO NOT CITE OR QUOTE
3-8
-------
capillary barrier by ET or percolation (breakthrough) into the coarser-grained soil layer. For
sloping capillary barriers (most common scenario), lateral diversion of infiltrating water provides
an additional means of removing soil water from the finer-grained soil layer. Lateral diversion is
essentially gravity-driven unsaturated drainage within the finer-grained layer. Because the water
content in the finer-grained layer is usually greatest near its interface with the underlying
coarser-grained soil layer, and the hydraulic conductivity of an unsaturated soil increases with
increasing water content, lateral diversion is concentrated near this interface. Laterally diverted
water causes the water content in the finer-grained soil to increase in the downdip direction. The
diversion length is the distance that water is diverted along the interface between the soil layers
before there is appreciable breakthrough into the coarser-grained layer. To avoid significant
breakthrough, the cover system slope length should be less than the diversion length (Figure 3-
5). Therefore, if a capillary barrier is sloped, the two-dimensional (lateral and vertical) effects of
soil-water movement must be taken into account in design of the barrier.
Ol ,".".".*.•.'.•.
"'- .-::::::::::::::::::::: Fine-Grained Soil
.•:•:::::::::::::::::::::::::::::::::::::::::::::: :;•:•.•!•!•!•!•! Coarse-Grained Soil
: 1 ric reia si hg: Wetter::::::::::: i: i: i:::':: '.•'.•'.•'.•'.•'.•'.•'.•'.•'.•'•'
Waste
Breakthrough
DL = Diversion Length
SL = Slope Length
L = Lateral Drainage
Figure 3-5. Problem Where Diversion Length is Less than Cover Slope Length on a
Capillary Barrier.
DRAFT - DO NOT CITE OR QUOTE
3-9
-------
Some advantages of incorporating a capillary barrier rather than an ET barrier alone in a cover
system include:
• The finer-grained soil layer of a capillary barrier stores more water than a comparable
layer without the capillary break (i.e., a free-draining layer). Compared to an ET barrier,
the additional storage capacity either serves to reduce overall percolation, or reduce the
total thickness required for the finer-grained soil to yield the same degree of percolation
inhibition.
• The additional water stored within a capillary barrier tends to encourage the
establishment and development of the surface vegetation. The increased vegetative
cover, in turn, removes more soil water due to greater ET. Furthermore, plants serve an
important function in reducing surface erosion.
• In addition to providing the capillary break, the coarser-grained layer of the capillary
barrier can serve as a biointrusion barrier and/or possibly a gas collection layer if small
amounts of gas are generated. (If gas emissions through the cover system are a concern,
gas containment components should be incorporated into the cover system design.)
Potential disadvantages of a capillary barrier compared to an ET barrier include the need to
specify and construct two different material types, the potential difficulties in constructing the
interface between the different materials (to form the capillary break), and minimizing
differential settlement.
3.3.2 General Issues
A number of the same general issues that were mentioned in Sections 2.2.1 and 2.3.1 for surface
and protection layers, respectively, also apply to the capillary barrier. Important issues are water
storage capacity and erosion potential, since excessive erosion can cause the cover to be
ineffective. In addition, it is particularly important to construct smooth and unmixed interfaces
between adjacent soil layers, as discussed in Section 3.5.2. Good CQA/CQC of these interfaces
is essential.
Two issues specific to capillary barriers were described by Koerner and Daniel (1997) and are as
follows: (i) the finer-grained soil must not be allowed to migrate over time into the underlying
coarser-grained soil; and (ii) over periods of extremely high precipitation, the capillary barrier
may cease to function, at least temporarily, as the coarser-grained soil becomes moist and more
permeable than the finer-grained soil. The former issue is discussed in more detail in Section
3.3.6. The latter issue is addressed by incorporating an appropriate factor of safety in design, as
discussed in Section 3.3.4.
3.3.3 Elements of Design
Important questions to be addressed when designing a capillary barrier include the following:
• How should the barrier be sloped?
• What materials should be used to construct the barrier?
DRAFT - DO NOT CITE OR QUOTE
3-10
-------
• How thick should the different layers be to store the required amount of water, wick
away infiltrating water, and create a capillary break?
• What surface treatments should be applied to control erosion?
• Which plants should be established to promote transpiration and stabilize the cover
surface?
• How and at what frequency should the barrier be maintained?
• What type and frequency of monitoring should be employed?
3.3.4 Design Concept
The design concept for the finer-grained soil component of the capillary barrier is essentially the
same as that presented for the ET barrier in Sections 3.2.4 and 3.2.5. The required minimum
thickness, however, can be less for a non-sloped capillary barrier than for an ET barrier. In
general, the capillary barrier increases the apparent field capacity of the finer-grained soil
component, thereby increasing the water storage capacity of this component. Consequently, the
finer-grained soil layer in a capillary barrier may not need to be as thick as the same layer used
alone in an ET barrier. In fact, the non-sloped capillary barrier may be preferred if the finer-
grained soil layer is required to be relatively thick. If this layer is too thick, all of the stored
water may not be removed by subsequent ET.
The apparent field capacity, 9afc, of the finer-grained soil component of a capillary barrier can be
estimated using a measured or modeled water content at which drainage from the capillary
barrier occurs (Stormont and Morris, 1998). This water content is greater than the soil's field
capacity due to the effects of the capillary break and can be calculated as:
(Eq.3.1)
where: 9 = volumetric water content; L = thickness of the finer-grained soil layer; z = distance
above the finer-coarser interface; and hz = minimum head at which flow into the coarser-grained
layer first occurs.
The texture of the finer-grained soil is important in determining the additional water storage
capacity achieved with a capillary barrier. Stormont (1996) described a field-scale (14 m2
surface area) water balance experiment conducted to measure the water storage capacity of a
capillary barrier. The barrier was comprised of a 900-mm thick layer of silty sand placed over
uniform gravel (0.6 mm). The barrier was installed at a 10% grade. The water content in the
finer layer, measured as added water, was increased at a constant rate of about 10 mm/day.
Breakthrough into the coarser layer was detected by collecting water that drained from the
coarser layer. The volumetric water content in the finer-grained layer at breakthrough was about
0.40 near its interface with the coarser-grained layer. Stormont (1996) estimated the total
amount of water stored in the capillary barrier at breakthrough by integrating the measured water
content over the thickness of the finer-grained layer. Expressed as a normalized quantity with
respect to area (volume of water divided by surface area), the capillary barrier stored 285 mm of
DRAFT - DO NOT CITE OR QUOTE
3-11
-------
water at breakthrough, which corresponds to an average apparent field capacity of approximately
0.32. The storage capacity of the capillary barrier can be compared to that estimated for a simple
ET barrier. Without the capillary break, water will drain approximately to the soil's field
capacity. The field capacity for the same soil (silty sand) can be estimated at 0.19, based on the
data for representative soils presented in Table 2-6. By integrating this water content over the
same 900 mm thickness, the silty sand in an ET barrier configuration would be expected to store
about 170 mm of water before drainage commenced. Thus, an additional 115 mm of water
storage was gained by the capillary break for the same cover soil thickness. In other words, a
simple ET barrier would need to be about 1510 mm thick to store the same amount of water as
900 mm of the same soil in the considered capillary barrier configuration.
The texture of the coarser-grained soil is also important in assessing the water storage capacity
of a capillary barrier (Khire et al., 2000). For example, if the coarser soil becomes more broadly
graded, hz in Eqn. 3-1 will decrease and 9afc will decrease. In contrast, if coarser soil becomes
more uniformly graded or if the average particle size of the coarser soil is reduced, hz* will
increase and 9afc will increase.
The design of a sloped capillary barrier also includes the selection of the slope gradient and the
distance between lateral drainage outlets to minimize the percolation of water through the
coarser-grained soil. These parameters can be assessed using a two-dimensional or three-
dimensional unsaturated flow computer model, such as HYDRUS-2D or VS2D-T. These models
are briefly described in Chapter 4. In general, layer thickness, diversion length, and slope
gradient requirements depend on climatological information for the specific site (e.g.,
precipitation, temperature, humidity) and the characteristics of the soils used in the cover (e.g.,
water storage capacity, hydraulic conductivity, texture). Other factors that should be taken into
consideration include slope stability, vegetation characteristics, and potential for desiccation
(Dwyer, 1997).
The lateral diversion capacity of the finer-grained layer is dependent in large part on the
hydraulic conductivity of the layer. In general, the hydraulic conductivities of silts and loams
are too low to permit appreciable lateral diversion. Field tests of capillary barriers with
homogeneous finer-grained layers indicate that the effective diversion lengths are less than 10m
(Nyhan et al., 1990; Hakonson et al., 1994; Stormont, 1995; Stormont, 1996; Nyhan et al., 1997).
These short diversion lengths are a consequence of the relatively low hydraulic conductivity of
the finer-grained soils compared to the infiltration rate during stressful periods when the soil is
relatively wet (e.g., spring snowmelt). Thus, soils that are often preferred as a rooting medium
and for their water storage capacity (e.g., loams, silts) may not be conductive enough to
substantially divert soil water laterally.
Utilizing "transport layers" or "unsaturated drainage layers" within the finer-grained layer
(Stormont, 1995) that allow water to drain laterally and outlet (e.g., in a swale) can increase the
diversion capacity of capillary barriers. Transport layers are one or more relatively conductive
layer(s) that drain water laterally within the cover's finer soil layer while remaining unsaturated.
Because soil water tends to accumulate near the interface between the finer and coarser layers
and unsaturated hydraulic conductivity increases with water content, a transport layer near the
interface is most effective in laterally diverting water. An effective transport layer, for example,
DRAFT - DO NOT CITE OR QUOTE
3-12
-------
could consist of a 300-mm thick relatively fine-grained, uniform sand that has a relatively high
hydraulic conductivity under moderate to high matric potentials. The lateral diversion afforded
by a transport layer complements the water storage function of the overlying soil, expanding the
conditions and climate for which a capillary barrier could be effective.
3.3.5 Coarser-Grained Soil Layer
The primary function of the coarser layer is to form a capillary break, but it may also serve as a
biointrusion barrier or, possibly, a gas collection layer.
Capillary break - The movement of water from the overlying finer-grained layer into the
underlying coarser-grained layer is controlled by the water entry potential of the coarser-grained
layer. The water entry potential is the potential associated with the movement of water into the
smallest pores that form a continuous network. Water will not move from an overlying moist
layer into an initially dry underlying layer at potentials less than the water entry potential suction
of the underlying layer. Using a coarser-grained soil with a higher water entry potential delays
the movement of water from the finer-grained soil layer into the coarser layer, permitting more
water to be stored in the finer layer near the interface (Figure 1-5). The suction head
corresponding to the water entry potential can be roughly approximated by the height of
capillary rise within a soil (Hillel and Baker, 1988). Thus, the water entry potential is expected
to be high for a uniform coarse-grained soil and decrease as the amount of fines in the soil
increase.
Biointrusion Barrier - As discussed in Sections 2.3.2.2.4 and 2.3.2.2.5, plants and animals
penetrating the cover system can create conduits for water to move downward into the waste,
and may even transport waste to the surface. Plant roots will generally not grow in soils with
water contents below the wilting point. Because coarse materials drain to low water contents,
typically below the wilting point, they can serve as barriers to root penetration. To be effective
as a root barrier, fines must be kept out of the coarse soil layer. This suggests that the particle-
size of the coarse layer material either has to be fine enough such that the overlying fines do not
penetrate into it, or an intermediate layer or a geotextile (GT) must be used to retain the
overlying soil, as discussed in Section 3.3.6. One design approach deterring animal invasion is
to use cobble-size particles that are too heavy for the animals to displace, as discussed in Section
2.3.2.2.5. Another approach is to use a dry, cohesionless uniform material that does not form a
stable burrow or tunnel.
Gas Collection Layer - For wastes that produce gas, it may be necessary to collect, transmit, and
potentially treat this gas as it is emitted from the buried waste. The coarser layer of the capillary
barrier may potentially be used for gas collection and transmission. If the facility is a landfill
subject to EPA's gas collection and treatment regulations or if gas emissions through the cover
system are a concern, the cover system should incorporate a gas collection system. While these
alternative designs may be adequate for hydraulic control, they should generally not be used
without gas containment components at MSW landfill sites where landfill gas collection and
treatment is required.
DRAFT - DO NOT CITE OR QUOTE
3-13
-------
3.3.6 Internal Stability
In general, the greater the contrast in texture or particle-size distribution of the fine and coarse-
grained soil components of a capillary barrier, the greater the effectiveness of the capillary break
(Stormont, 1997). There is concern, however, that finer soil particles will move into the pores of
the coarser soil, degrading the interface and reducing the effectiveness of the capillary break.
The conventional approach for evaluating the internal stability of the capillary barrier is to
ensure the soils satisfy a soil retention criterion. The retention criterion establishes the
relationship of grain sizes of adjacent materials necessary for the coarser material to retain the
finer material. The retention criteria for soil and geotextile filters are discussed in Section 4.7.
From conventional filter criteria, interface stability is favored by soils having similar particle-
size distributions, apparently in conflict with maximizing the effectiveness of a capillary break.
Conventional criteria, however, have been developed using high hydraulic gradients for
applications such as dams. In contrast, capillary barriers would only rarely, if ever, experience
positive pore pressures, and the associated hydraulic gradients would be small. Furthermore,
capillary barriers will be subjected to cycles of wetting and drying in response to climatic
conditions. Thus, interface stability should be considered under dry conditions, as well as, under
relatively small positive water pressures. The biggest risk to internal stability of a capillary
barrier may occur during barrier construction. For example, vibratory compaction could cause a
large number of fine particles to move into the coarser layer.
Koerner and Daniel (1997) recommend that a GT separator be considered at the capillary barrier
interface. They indicate that for extremely long service times (e.g., hundreds of years) fiberglass
GTs have been considered for this application. It is noted, however, that with a GT at the
capillary barrier interface, the capillary break may occur between the finer-grained soil and GT
rather than between the finer- and coarser-grained soils (Stormont et al., 1997). This effect
reduces the water storage capacity of the finer-grained soil. The GT could also function as a
lateral drainage layer. If it is necessary to use a GT separator, the effects (reduced water storage
capacity and lateral drainage) associated with use of the GT should be considered and addressed
in the final capillary barrier design.
3.4 Alternate Design Performance Evaluation
3.4.1 Introduction
The preceding sections highlighted how the water storage and lateral diversion characteristics of
ET and capillary barriers are affected by factors such as soil type and thickness and slope of the
interface. In addition to the influence of material properties and configuration, the "stress"
provided by the climate will have a major impact on the performance of these types of barriers.
To accommodate these factors into the development of designs and estimating the performance
of ET and capillary barriers, numerical simulations can be used. However, numerical
simulations have two challenging aspects that must be addressed to enable reasonable
representation of actual field conditions. First, for near-surface applications it is necessary to
account for the effect of time- and climate-dependent processes, including precipitation, soil
water evaporation, and plant transpiration. The second aspect, specific to capillary barriers, is
that water movement within the near-surface soils and near the interface is transient, unsaturated
DRAFT - DO NOT CITE OR QUOTE
3-14
-------
flow involving materials of widely varying properties. Accuracy and stability of numerical
solutions involving these types of flow behavior can be difficult to achieve.
As previously discussed in Section 1.2.3, EPA recommends that a cover system be designed to
minimize percolation to prevent the bathtub effect, with a specific value selected based on the
nature of the contained waste, the hydrogeological vulnerability of the site, and other factors.
The Agency considers this performance criterion to apply over a considered performance period
(e.g., maximum rate over at least a 30-year post-closure simulation).
Numerical modeling should be used to design a cover system that meets this performance
criterion. Natural analogs may be used to help predict long-term cover performance, and field
monitoring may be required, depending on site-specific percolation criteria.
3.4.2 Numerical Modeling
Computer numerical simulations can be used to predict the water balance performance of a cover
system. Computer simulations are only as good as the input data provided and the system
modeled. Much of the difficulty comes in obtaining good and accurate input data to correctly
predict a cover system's water balance performance. It is advised that a realistic set of input
parameters be developed for the simulations based on measurements from the actual soil to be
used (at the anticipated installed density and moisture content), values from the literature, and
expert opinion. Generally, input properties include unsaturated soil properties (i.e., moisture
characteristic curves - matric potential versus moisture content) and hydraulic conductivity.
There are a number of practitioners who believe that even a near perfect set of input data and a
well-designed computer model will still not yield reliable results. Because of this limitation, it
may be prudent in critical applications to not rely solely on the results of one set of computer
model predictions and/or to use a larger factor of safety. It is suggested that for critical
applications, two different computer models be employed and the results of the simulations
compared.
The EPA HELP computer model (Schroeder et al., 1994a,b) is at present the industry standard
for conducting water balance analyses for conventional hydraulic-barrier cover systems. This
model is discussed in more detail in Section 4.2.3.2. Field applications of the model are
discussed in Section 4.3. The HELP model incorporates a number of simplifying assumptions
and does not solve the unsaturated flow equations. Thus, it is not considered particularly good
for evaluating ET barriers and it is not recommended for evaluating capillary barriers.
Unfortunately, there are no public-domain water balance models currently available that are as
user friendly as HELP and that properly model unsaturated flow within the cover system soil
layers.
A model that may be used for the analysis of ET and capillary barriers is UNSAT-H (Payer and
Jones, 1990), a one-dimensional finite-difference computer program to solve for water and heat
flow in soils. This model is discussed in more detail in Section 4.2.8. Field applications of the
model are discussed in Section 4.3. The UNSAT-H code solves Richard's partial differential
equation (Richards, 1931) and can be used to simulate the water balance for evapotranspirative
or non-sloped capillary barriers. However, the vegetation options in the model were developed
DRAFT - DO NOT CITE OR QUOTE
3-15
-------
for the DOE Hanford site near Richland, Washington and may not be applicable to other areas of
the country. The model user either assumes that: (i) the vegetation is similar to cheatgrass; or
(ii) vegetation quantity is based on a daily leaf area indices input by the user. The vegetation is
required to start germinating from a seed before Julian day 91 or after day 273 and to stop
transpiring between Julian days 151 to 243. In some areas of the southwest, Tucson, Arizona, for
instance, relatively high precipitation and plant transpiration is still occurring after Julian day
243.
Other models that may be considered and that are discussed in this guidance document are
LEACHM (Section 4.2.3.3.), SoilCover (Section 4.2.3.5), and, for sloping capillary barriers,
HYDRUS-2D (Section 4.2.3.6). All of the models have their specific advantages and
disadvantages, some of which are listed in Table 4-1.
3.4.3 Performance Monitoring
Because of design and construction quality control uncertainties, performance monitoring is
recommended for alternative covers. Field performance data provide perhaps the most reliable
information for assessing whether cover systems are performing as designed. It is recommended
that a project specific monitoring system be utilized to monitor the performance of an ET or
capillary barrier throughout the life of the cover system. As an example, a lysimeter used by the
ACAP program for monitoring landfill cover performance is shown in Figure 3-6. Additional
performance monitoring techniques are discussed in more detail in Chapter 8.
Examples of performance monitoring of alternative cover systems are highlighted below:
• Albuquerque, New Mexico. The Alternative Landfill Cover Demonstration (ALCD) is
a large-scale field test at Sandia National Laboratories located on Kirtland Air Force Base in
Albuquerque, New Mexico (Dwyer 1997, Dwyer 1998, Dwyer 2001). Six landfill cover
profiles are installed with automated retrieval of water balance data (runoff, lateral drainage,
percolation, soil moisture changes within the covers, and precipitation). The covers are
periodically stress tested by adding precipitation to the covers through sprinkler systems to
simulate worst case infiltration events at various locations in arid and semi-arid climates.
Four alternative covers (ET Cover, 2 different Capillary Barrier Designs, and a cover
featuring a GCL) are installed next to two prescriptive covers (RCRA Subtitle D - similar to
Figure l-6(a) and RCRA Subtitle C - similar to Figure 1-7) for direct water balance
performance comparison. The project's intent is to compare and document the performance
of alternative landfill cover technologies of various costs and complexities for interim
stabilization and/or final closure of landfills in arid and semi-arid environments. The test
covers are constructed side-by-side for comparison based on their performance, cost and ease
of construction. The ALCD is not intended to showcase any one particular cover system.
The focus of this project is to provide the necessary tools; i.e., cost, construction and
performance data so that design engineers can support less expensive, regulatory acceptable
alternatives to conventional cover designs. This project has been extensively reviewed by
regulators from across the country as well as by panels from the National Academy of
Science and the Department of Energy. Results from this project have shown properly
designed alternative covers such as ET Covers and Capillary Barriers are as good as or better
than their prescriptive counterparts. Results from this demonstration have been used by a
DRAFT - DO NOT CITE OR QUOTE
3-16
-------
number of regulatory agencies to approve permits for the use of an alternative landfill cover
in lieu of a prescriptive cover (Dwyer 2001).
• EPA's Alternative Cover Assessment Program (ACAP): (http://www.acap.dri.edu/)
• Sierra Blanca, Texas (http://www.beg.utexas.edu/environqlty/vadose/index.htm)
3.4.4 Natural Analogs
Conventional engineering approaches for designing landfill covers often fail to fully consider
ecological processes. Natural ecosystems effective at capturing and or redistributing materials in
the environment have evolved over millions of years. Consequently, when contaminants are
introduced into the environment, ecosystem processes begin to influence the distribution and
transport of these materials, just as they influence the distribution and transport of nutrients that
occur naturally in ecosystems (Hakonson et al., 1992). As described in Section 1.5.6, as the
ecological status of the cover changes, so will performance factors such as water infiltration,
water retention, ET, soil erosion, gas diffusion, and biointrusion (Caldwell and Reith, 1993). An
important objective for an effective cover system is to design it so that subsequent ecological
change will enhance and preserve system performance. Consideration of natural analogs can
enhance a cover system design by disclosing what properties are effective in a given
environment or what processes may lead to possible modes of failure. These factors can in turn
be avoided during the design and construction phases. Natural analog studies provide clues from
past environments as to possible long-term changes in engineered covers. Analog studies
involve the use of logical analogy to investigate natural and archaeological occurrences of
materials, conditions, or processes that are similar to those known or predicted to occur in some
part of the engineered cover system (Waugh, 1995).
One possible analog might be observed by trenching adjacent to the site in an undisturbed area
and measuring the depth of plant roots (Dwyer et al 1999). This will reveal the general depth of
infiltration. Another method for assessing the average long-term depth of water penetration (or
infiltration depth) is to trench adjacent to the site in an undisturbed area to observe the depth of
calcium carbonate (CaCOs) deposits or formation of a caliche layer. Soils in semiarid and arid
regions commonly have carbonate-rich horizons at some depth below the surface. The position
of the CaCOs bearing horizon is therefore, related to depth of leaching, which, in turn, is related
to climate (Birkeland, 1984).
DRAFT - DO NOT CITE OR QUOTE
3-17
-------
LLDPE
Cutoff
Fillet
Welded
to
Base
GM 0.3
\ —
V
r
k
K
Earthen /X
Berm ^_^^
^^vl
^ """>
_Dr£
(25
Sump Leak Cover
Test Pipe (thickness site
/specific)
ieosynthetic /
oot Barrier/
/
/ /
i
sinage Pipe
5 mm PVC)
//
7
J
\ '
— 5 — — T~
•
\ 1.5 mm LLDPE GM \ \
(textured both sides) \ \
fe.
Vadose
Zone
Monitoring
Stations
7
r
Berm
Interim
^nvpr 5-inil
LLDPE
Cutoff
Fillet
Welded
to
Base
,GM
PV
17
X
Compacted
Base
(thickness
varies)
Notes:
1. Base shall be approved fill (fines > 30%) compacted to >95% of max. dry unit weight based on ASTM D 698 and dry of optimum water content.
2. Smooth base before placing GM. Eliminate all ridges, depressions, etc. > 25 mm in height. Remove all stones, etc. larger than 10 mm.
3. Place GM in early mornining and ensure good contact with all surfaces. No gaps shall exist between base and GM.
4. Vertical cutoff sheets shall be fillet welded to base GM.
5. GC drain is GN with non-woven GT heat bonded to both sides. Install using rub sheet.
6. Interim cover soil shall be placed on GC drain from the edges. No equipment can travel on GM or GC drain. Once spread in 450 mm loose lift,
compact to >85% of max. dry unit weight based on ASTM D 698.
7. Vadose zone monitoring stations instrumented with TDRs to measure soil water content and heat dissipation sensors to monitor soil water pressure
and temperature.
Figure 3-6. Test Plot Design Used at ACAP Sites.
DRAFT - DO NOT CITE OR QUOTE
3-18
-------
The origin of carbonate horizons involves carbonate-bicarbonate equilibria (Birkeland, 1984), as
shown by the following reactions:
CO2 + H2O
CaCO3 + H2CO3 <-» Ca2+ + 2HCCV
s aq aq aq
Carbon dioxide partial pressures in soil air are 10 to more than 100 times that in the atmosphere;
this decreases the pH, which, in turn, increases CaCO3 solubility. The partial pressure of CO2 is
high as a result of CO2 produced by root and microorganism respiration and organic matter
decomposition. Thus, one would expect the highest CO2 partial pressure to be associated with
the A horizon located near the surface, with values diminishing down to the base of the zone of
roots. In arid and semi-arid regions, the quantity of water leaching through the soil is also
generally greater near the surface than at depth. Thus, as the water moves vertically through the
soil, the Ca+ and HCO3" content might increase to the point of saturation after which further
dissolution of CaCO3 is not possible. Combining the effects of high CO2 partial pressure and
downward-percolating water, the formation of CaCO3-rich horizons may be understood as
follows. In the upper zone of the soil, Ca2+ may already be present or may be derived by
weathering of calcium-bearing minerals. Due to plant growth and biological activity, CO2
partial pressure is high and forms HCO3" upon contact with water. Water leaching through the
profile carries Ca2+ and HCO3" downward in the profile. Precipitation of CaCO3 to form a
caliche horizon takes place by a combination of decreasing CO2 partial pressure below the zone
of rooting and major biological activity and the progressive increase in Ca2+ and HCO3"
concentrations with depth in the soil solution as the water percolates downward and water is lost
by evapotranspiration. The position (depth) of the CaCO3 bearing horizon is therefore related to
depth of leaching, which, in turn, is related to the climate.
As more alternative cover systems are installed and demonstrate successful performance,
confidence for their use at other sites will grow. A number of experiments and field-scale
demonstrations throughout the country are currently producing field data to document the short-
term performance of alternative cover technologies (Dwyer, 1997; Dwyer, 1999; Dwyer, 2001;
Benson, 1997). As with any emerging technology, longer-term performance data are lacking.
Natural analogs can be used to deduce how a system may perform over a longer period (Waugh,
1995). Computer modeling can be used to predict long-term performance and compare
alternative designs (Khire, 1995; Morris and Stormont, 1997a, b). Until long-term performance
data have been obtained, the combination of computer model predictions, field data, and natural
analog studies forms the basis for evaluating long-term alternative cover system designs.
DRAFT - DO NOT CITE OR QUOTE
3-19
-------
3.5 Construction of Alternative Designs
ET covers may often be easier to build and require a lesser amount of quality assurance
(QA)/QC during construction than conventional designs with hydraulic barriers. This is due to
the fact that the ET cover may only involve placement of two soil types, a topsoil layer and the
relatively fine-grained ET barrier, and no geosynthetics or soils that must be compacted to meet
strict hydraulic conductivity criteria. The complexity of construction of a capillary barrier
increases with the number of layers in the system, including layers for soil water storage, internal
drainage, biointrusion resistance, and/or gas transmission.
Specific construction and maintenance considerations for alternative cover system designs are
discussed below.
3.5.1 Compaction Requirements
CCL hydraulic barriers in conventional cover system designs are compacted to attain a very low
saturated hydraulic conductivity. As discussed in Section 2.5.4 of this document, this generally
requires compacting the soil lifts 'wet of optimum' to remold the soil and produce high soil
densities. Compacting the soil wet of optimum increases the potential for desiccation cracking
and reduces the initial water storage capacity since the CCL is generally at a degree of saturation
of at least 85%.
The alternative cover system designs outlined in this chapter are designed to function under
unsaturated conditions; consequently obtaining very low saturated hydraulic conductivity is not a
priority. Because a very low initial saturated hydraulic conductivity is not the objective when
placing finer-textured soils in an alternative cover system, compaction "dry of optimum" is
usually desired to reduce the potential for desiccation cracking. This compaction alternative also
allows for additional initial water storage capacity and a structure that is less restrictive to plant
roots. Compaction density requirements for the finer-grained soils should be based on
consideration of the water content-unsaturated hydraulic conductivity relationship for the soil,
erosion resistance, and plant rooting requirements. Generally, compaction for the ET barrier is
performed in an attempt to mimic the naturally occurring in-situ soil density for a particular
borrow material. Ideally, target densities for constructed ET cover soils should be within +/- 5%
of the in-situ borrow soil density. In addition, this target in-situ soil density should be used for
any subsequent laboratory testing and for input parameters in computer water balance models. It
should be noted that unsaturated soil properties and saturated hydraulic conductivity are very
sensitive to the soil's density. Uniformity of compaction is critical.
3.5.2 Capillary Barrier Soil Interfaces
During the emplacement of a capillary barrier, special care must be taken during the placement
and compaction of the first lift of fine-grained soil on the underlying uncompacted coarse-
grained soil. The interface between these two materials should remain smooth and continuous
and the materials should not be mixed together.
Heavy compaction, especially if a vibratory compactor is used, should be avoided as finer soil
may migrate into the coarser layer. Conversely, a lack of compaction will leave the finer soil
DRAFT - DO NOT CITE OR QUOTE
3-20
-------
near the interface in a loose condition. This finer soil could be more prone to internal erosion
under the action of seepage forces should gravity-driven water percolation develop at the
interface. Small wide-tracked bulldozers have been used to construct this interface. The steel
tracks help distribute the weight of the bulldozer over a greater surface area, thus reducing its
contact pressure. Kneading compaction is not recommended for the first lift of fine-grained soil;
rather a smooth drum roller should be used. This will help minimize the potential for mixing of
fine and coarse soils at their interface. The design process for capillary barriers should include
an evaluation of appropriate procedures for soil compaction.
3.6 Maintenance and Monitoring of Alternative Designs
3.6.1 Maintenance
Maintenance is discussed in Chapter 9. The most important maintenance activities for the
alternative designs involve maintaining the intended vegetative cover and the erosion control
measures, repairing erosion gullies, surface depressions caused by localized settlement, surface
cracks, and, as an associated activity, maintaining and repairing surface-water management
structures.
Maintaining the surface layer and repairing cracks and erosion gullies in alternative cover
systems is generally even more critical than maintaining the surface and protection layers in
conventional cover systems that have a drainage layer and a GM barrier. A crack in an
alternative cover system may allow short circuiting of water through the cover system and impair
cover system performance. If differential settlement of an ET barrier occurs, the barrier can
simply be repaired by applying more soil to the surface to bring the cover system back to its
original grade. For a capillary barrier, the repair is more complex. The finer-grained soil first
should be excavated to expose the coarser-grained soil, and the depression in the coarser-grained
soil should be filled with the coarser soil so that the interface between the finer and coarser-
grained soils is brought back in-line with that adjacent to it. The finer-grained soil at the repair
location should then be blended in with (e.g., stair-stepped into) the surrounding finer-grained
soil to reduce the potential for preferential pathways for infiltrating water.
3.6.2 Monitoring
Monitoring is discussed in Chapter 9. Alternative cover systems should be monitored to identify
problems with excessive erosion, excessive differential settlement, excessive cracking, or slope
instability, assess the health of the vegetative cover, and evaluate gas emissions, if gases are a
concern. If the cover system water balance is being assessed, the soil moisture content or matric
potential, percolation through the cover system, and surface-water runoff may also be monitored.
DRAFT - DO NOT CITE OR QUOTE
3-21
-------
3.7 Alternative Materials
3.7.1 Geofoam
As described by Horvath (1995a), geofoam refers to any manufactured material created by some
internal expansion process that results in a material with a texture of numerous, closed, gas-filled
cells. The cell walls are solid, although generally relatively thin and permeable to gases.
Currently, the most common geofoam material is expanded polystyrene (EPS), a white foam that
is also used for non-geofoam applications, like beverage cups and packaging. It is noted that
EPS, along with extruded polystyrene (XPS), another geofoam material, are both referred to by
ASTM as rigid cellular polystrene (RCPS) in below-grade applications (Horvath, 1995a). This
lightweight material of a density between 10 and 20 kg/m3 has unique engineering properties.
White (1995) presents the following data as typical of EPS:
• water absorption is very low, e.g., 2% (maximum) by volume;
• low temperatures, under-water or wet environments, and exposure to freeze-thaw cycling
do not adversely impact mechanical properties;
• EPS is a very efficient thermal insulator (because it is approximately 98 to 99% gas by
volume), and this feature has been capitalized upon in several landfill applications; and
• the mechanical properties of elastic modulus, Poisson's ratio, and compressive strength
are readily assessed by either static or cyclic loading tests.
According to Horvath (1995a), the only concern with using EPS and XPS geofoams is that they
may degrade when in contact with certain chemicals (i.e., petroleum hydrocarbons and, possibly
the plasticizer in PVC GMs.
Geofoam has been used above the drainage layer and barrier of a cover system for insulation and
because of its lightweight properties (Gasper, 1990). It has also been used as a spray for daily
landfill cover (Gasper, 1990), beneath a GM as a smooth protection layer over steep slopes in an
abandoned quarry (Horvath, 1995b), and to promote methane and radon gas venting (White,
1995 b).
3.7.2 Shredded Tires
Scrap automobile and truck tires represent a large quantity of waste material that can be used in
select construction, operations, and closure applications for waste containment facilities. In the
U.S., an estimated 280 million scrap tires are generated annually. When cut into pieces,
typically ranging from 50 to 300 mm in length, shredded tires may be used in cover systems as
the gas collection layer, the drainage layer, the protection layer, or a component of the
foundation layer (GeoSyntec Consultants, 1998a,b,c).
Modern tires are composed of a combination of natural rubber and synthetic rubber elastomers
derived from oil and gas. Multiple carbon blacks, extender oils, waxes, antioxidants, and other
materials are added to enhance performance characteristics and manufacturing efficiency. Tires
contain a bundle of high tensile strength steel wires surrounded by rubber that forms the bead of
a tire to provide a firm contact with the rim. The individual wires that compose this bundle can
DRAFT - DO NOT CITE OR QUOTE
3-22
-------
be up to 3 mm in diameter and are relatively stiff. Most tires also contain steel belt wire in the
tread and sidewall areas. This wire is much smaller diameter than bead wire and is therefore
more flexible.
Metal wires protruding from tire shreds may scratch or puncture GMs and GCLs used in a cover
system. Therefore, whenever tire shreds are used in a cover system, careful consideration should
be given to the design of adequate protection (e.g., a geotextile or a soil layer between the tire
chips and GM) for the cover system geosynthetics. To minimize the potential for bead wires to
puncture a GM or GCL, the bead wire protrusions from the tire shreds should be limited (to less
than 10 mm for example) and a GT or soil cushion layer should be considered. Project-specific
laboratory or field testing is recommended. Tire shreds containing bead wire should not be
placed in contact with geosynthetics: either the bead wire needs to be removed or a soil layer
needs to be placed between the tire chips and the geosynthetics. Belt wire can also be
problematic. The results of a field test program (GeoSyntec, 1998b) show that belt wires in
direct contact with a GM can create some minor damage (i.e., indentations, scratches, dents). To
reduce the potential for GM damage by protruding or loose belt wire, the GM should be
separated from the tire chips by a GT or soil layer. The wires exposed at the cut edges of tire
shreds can also be a hazard to personnel walking on the shreds, and can puncture the tires of
vehicles trafficking over them. Track mounted or steel-wheeled equipment should be used when
practical to mitigate the latter problem.
The exposed metal in tire shreds may also leach metals when exposed to water; however, with
exceptions of iron and manganese, the metal concentrations are anticipated to be below their
primary or secondary drinking water standards (Duffy, 1996; Humphrey et al., 1997).
Tire shreds are combustible at temperatures above 322 °C. Combustion generally requires an
external ignition source (e.g., lightning), although there have been several fires in tire-shred fills
used for highway embankment fills that seem to be associated with spontaneous combustion due
to self-heating. Humphrey (1996) describes three fires that occurred during 1995 in tire shred
fills that were at least 6 m deep. Two of these fills are located in Washington, and one is located
in Colorado. Humphrey gave several potential mechanisms for ignition of the tire shreds, with
the most likely mechanism being oxidation of exposed steel wires. To reduce the potential for
future tire fires, Humphrey recommends minimizing the amount of steel belt exposed at the cut
edges of tire shreds, minimizing the amount of crumb rubber in the shred material, covering the
shreds with at least 1.2 m of soil to limit contact of the shreds with oxygen, not placing organic
materials (e.g., topsoil) directly over the shreds, and preventing contact between the shreds and
fertilizer. These recommendations may be appropriate for relatively deep fills, but appear to be
very conservative for applications where tire shreds are used in a cover system drainage layer or
gas collection layer.
Physical characteristics of tire shreds are dependent upon the shred size (gradation), uniformity,
exposed wire content, and whether the shreds have been mixed with soils. Compared to natural
materials (i.e., sands and gravels) typically used as drainage layer materials, tire shreds have a
much larger size. If the tire shreds are used as a drainage or gas collection material, soil or GT
filter or separation layers are often required between the shreds and the adjacent materials.
DRAFT - DO NOT CITE OR QUOTE
3-23
-------
Based on data complied from Ahmed (1993), Humphrey et al. (1993), and Cecich et al. (1996),
loosely dumped tire shreds typically exhibit dry densities between 4.0 and 4.8 kN/m3; the density
of compacted tire shreds is typically between 5.5 and 7.5 kN/m3.
Tire shreds are relatively compressible. Laboratory tests on compacted tire shreds less than 75
mm in length indicate that tire shreds may exhibit vertical strains of up to approximately 20%
under low vertical stresses up to approximately 25 kPa (Ahmed, 1993; Nickels, 1995). Tire
shred compressibility under the anticipated overburden stress should be accounted for when
specifying the minimum thickness of the as-compacted tire shred layer. Because they are so
compressible, construction of CCLs over tire shreds may be difficult. GeoSyntec Consultants
(1998a) showed that construction of a CCL directly over 300 mm of foundation soil underlain by
300 mm of tire shreds resulted in the development of numerous cracks in the CCL as the tire
shreds compressed. Such a relatively thin soil layer over the tire chips made it difficult to obtain
the required compaction density in the overlying CCL. Additionally, the weight of a sheepsfoot
roller or similar equipment used for compaction of a CCL could cause deflections of the tire
shreds in the foundation layer that would be large enough to introduce cracks into the CCL,
thereby increasing its hydraulic conductivity. When the foundation layer was modified to a 450-
mm thick soil layer over a 150-mm thick tire shred layer, the foundation was adequate for
construction of a CCL with a hydraulic conductivity of 1 x 10"6 cm/s or less. These results are
dependent on the size of the tire shreds and the thickness of the tire shred layer. All other things
being equal, smaller shred size and a thinner shred layer will provide more constructible
conditions than if these parameters were reversed. A field test program may be considered when
assessing the feasibility of constructing a CCL on top of a tire chip layer. The compressibility of
tire shreds may also preclude placement of GM directly over a tire shred layer. This is mostly a
problem during construction when construction equipment imposes stresses on the GM. For
example, the deformation imposed by a low-ground pressure dozer spreading a 0.3-m thick soil
layer over the GM may be sufficient to tear welded seams. Moreover, the compressibility of the
shreds directly under the GM may complicate placement of the GM itself (i.e., it may be difficult
to unroll the GM and the weight of field personnel may cause deformations that are sufficient to
complicate field welding). In the absence of a field test program to investigate this issue,
GeoSyntec Consultants (1998b) has recommended that at least 0.3 m of soils be placed over the
tire shreds to allow construction of the GM and overlying soil layers.
When comprising a gas collection or drainage layer, tire shreds must be able to provide the
required flow capacity under the applied normal stress. This is typically not a problem given the
relatively low stresses in cover system applications; however, at higher normal stresses, tire
shred compression and hydraulic conductivity reduction may be significant. Various tests have
indicated the hydraulic conductivity of 12 to 75 mm long tire shreds to be on the order of 0.006
to 0.79 m/s (Edil et al., 1992; Glade et al., 1993; Duffy, 1996) under relatively low normal
stresses. The lower end of this range corresponds to smaller tire shreds. High variability in
hydraulic conductivity values are due to differences in shred size, initial density, hydraulic
gradients, and confining pressures under study conditions.
Available published data on the shear strength of tire shreds indicate a wide range of shear
strength properties for tire shreds and tire shred/soil mixtures. The data are from varying test
types and test conditions. Humphrey et al. (1993) present data from large-scale direct shear tests
DRAFT - DO NOT CITE OR QUOTE
3-24
-------
conducted on tire chips with three different gradations. At normal stresses ranging from 14 to 68
kPa, the reported failure envelopes (i.e., friction angle and cohesion intercept) ranged from 19°
and 11.5 kPa to 26° and 4.3kPa. At the lower end of the normal stress range (i.e., 14 to 17 kPa),
these measured shear strengths yield equivalent secant friction angles of 38 to 45°.
3.7.3 Sprayed Elastomers
Although sprayed elastomers, such as polyurethane and polyurea, have been used for
waterproofing secondary containment systems, concrete water tanks, tunnels, roofs, and other
structures, there has been limited application of these materials to waste containment or
remediation sites. Sprayed elastomers could potentially function as gas and/or hydraulic barriers
in cover systems at these sites. These materials are typically easier and faster to apply than other
cover system barriers materials. Sprayed elastomer barriers have fewer seams than GM barriers.
However, these materials have not yet been used in a full-scale cover system application, and
the installation quality control and quality assurance procedures for such an application are still
being developed.
An elastomer barrier can be installed by heating an elastomer, pressurizing it, and spraying it
onto a surface. The material can be applied directly to a prepared soil subgrade. However, it
may be difficult to achieve a continuous barrier with a uniform finish using this installation
practice, especially if the subgrade surface has cracks. Therefore, in a cover system barrier
application it may be more appropriate to spray the elastomer onto a lightweight nonwoven
heatbonded GT placed without wrinkles or folds on a soil subgrade.
Laboratory testing has been conducted on factory-sprayed and field-sprayed polyurea elastomer
samples. Factory-sprayed samples were obtained from the material supplier, and field-sprayed
samples were collected from a30mx30m test plot installed in 1993 at a landfill in Michigan.
As described by Miller et al. (1997), the test plot included subplots with elastomer sprayed over
a prepared soil subgrade with some cracks, over a moist prepared soil subgrade with less cracks,
over a lightweight nonwoven heatbonded GT placed on a prepared soil subgrade, and over a
woven GT placed on a prepared soil subgrade. Half of the sprayed area on each subplot was
covered with approximately 150 mm of soil and the other half was left exposed. The results of
mechanical and hydraulic tests conducted on the factory-sprayed elastomer samples and
interface direct shear tests conducted on the field-sprayed elastomer samples are presented by
Cheng et al. (1994). According to Miller et al. (1997), the elastomer sprayed over the nonwoven
GT appeared to provide the best barrier installation. Field samples have been removed from this
barrier at periodic intervals to assess long-term performance. No significant degradation or
deterioration in the mechanical or hydraulic properties of the barrier samples has been observed
(Miller et al., 1997).
It should be recognized that sprayed elastomers have not yet been used in a full-scale cover
system application. While this type of application holds some promise, additional research and
development is necessary.
DRAFT - DO NOT CITE OR QUOTE
3-25
-------
3.7.4 Paper Mill Sludges
Paper mill sludges have been shown to have properties similar to those of clays and, as a
consequence, have been used as the hydraulic barrier material for some landfill cover systems in,
at least, Maine, Wisconsin, and Massachusetts (Zimmie and Moo-Young, 1995). From the
limited engineering properties data available for paper mill sludges, the properties vary
considerably among the sludges depending on the manufacturing process, water content, organic
content, sludge age, degree of consolidation, and other factors. Since the sludges are degradable,
their properties are time dependent. The degradation processes also generate gases, which must
be managed.
Zimmie and Moo-Young (1995) performed laboratory tests to evaluate the water content,
organic content, specific gravity, permeability, compaction, consolidation, and strength
characteristics of seven paper mill sludges of various ages. They found that the sludges had a
high initial water content ranging from 150 to 268%, an initial solids content of 27 to 40%, and
an initial hydraulic conductivity ranging from about 5 x 10"10 to 5 x 10"8 m/s, and behaved
similarly to highly organic soil.
Zimmie and Moo-Young (1995) also performed laboratory tests on six undisturbed samples of a
sludge used as the cover system barrier material for a MSW landfill in Massachusetts. Three
samples of the sludge were obtained shortly after construction and the other three samples were
taken at 9, 18, and 24 months after construction. The results of the laboratory tests on these
samples indicated that the water content and hydraulic conductivity of the sludge decreased
somewhat over time, presumably as the sludge consolidated and biodegraded (i.e., it mineralized
to become more like a kaolin clay).
Moo-Young and Zimmie (1996) evaluated how freeze-thaw affects the hydraulic conductivity of
paper mill sludges through a series of laboratory tests on sludge samples and by monitoring the
depth of frost penetration in the sludge barrier for the previously-mentioned MSW landfill in
Massachusetts. Based on the results of their laboratory tests, performed over a range of water
contents, if a sludge barrier is subjected to freezing and thawing cycles, the hydraulic
conductivity of the sludge may increase by one to two orders of magnitude. Over the several
year field study, the frost layer had not penetrated into the sludge barrier due to the protection
provided by the overlying soil layers and the high water content of the sludge.
When using paper mill sludge in a cover system application, the chemical characteristics of the
sludge need to be considered. Water percolating through the sludge may mobilize volatile
organic compounds and heavy metals contained in the sludge. To keep certain chemicals from
leaving the site (e.g., as runoff), paper mill sludge may be required to be the barrier or be located
below the barrier. Depending on its chemical properties, it may not be suitable for use as a
protection layer.
DRAFT - DO NOT CITE OR QUOTE
3-26
-------
Chapter 4
Hydraulic Analysis and Design
4.1 Introduction
This chapter provides information on select topics related to cover system hydraulic analysis and
design. The specific topics discussed in this chapter are:
• characteristics of selected water balance models (Section 4.2);
• evaluation of the water balance models (Section 4.3);
• recommendations for application of the water balance models (Section 4.4);
• design of drainage layers (Section 4.5);
• design of slope transitions (Section 4.6); and
• design of filter layers (Section 4.7).
4.2 Characteristics of Water Balance Models
4.2.1 Overview
As described in Section 1.2.5, with EPA's liquids management strategy, a primary function of a
cover system is to limit post-closure leachate generation by minimizing or preventing, for all
practical purposes, percolation of water into the waste. A water balance analysis is used to
predict the quantity of this percolation. In addition to estimating percolation, water balance
analyses of cover systems are used to:
• develop an understanding of how the various cover system components will function and
identify which water routing mechanisms are most important;
• compare the performance of different cover system designs; and
• define the performance criteria for various cover system components (e.g., required
storage capacity of surface and protection soil layers, required flow capacity of drainage
layer) so that these components can be designed.
This section of the guidance document describes the water balance concept and presents several
water balance analysis methods commonly used for cover systems.
4.2.2 Water Balance Concept
In a water balance analysis, water is routed into and out of a system using a series of calculations
that require conservation of water mass. The potential pathways for water movement into and
out of a cover system are illustrated in Figure 4-1. A cover system water balance is expressed in
DRAFT - DO NOT CITE OR QUOTE 4-1
-------
terms of water inflows and outflows and storage changes for a unit area of the system over some
arbitrary time interval as:
P = R + ET + AWsurface + AWsoil + L + PERC + AWfoHage
(Eq.4.1)
where: P = precipitation (mm/day); R = runoff (mm/day); ET = evapotranspiration (mm/day);
AWsurfaCe = change in water storage at surface (mm/day); AWf0iiage = change in water storage on
plant foliage (mm/day); AWsoii = change in water storage in cover system soil (mm/day); L =
lateral drainage (mm/day); and PERC = percolation through the cover system (mm/day). Water
is input to the cover system as precipitation in the form of rain or snow and lost from the cover
system by runoff, ET, lateral drainage, and percolation. Water also is stored on the cover system
as ponded water or snow, on plant foliage, and in cover system soils by capillary action. Eq. 4.1
is cast above using a time interval of one day; the equation could be developed using any other
time unit.
Storage as Snow
Storage on Foliage
Surface/
Protection Layer
Drainage
Layer
Hydraulic
Barrier
Foundation
Layer
Precipitation
I
Runoff
Infiltration' '•'•-'••'= • V '/....
;. . . . Evapotranspiration
Percolation
——__ _ Waste or-_—_——_ ~^_
—~-_^_ ~^— Contaminated — _^r-
Material
Figure 4-1. Water movement and storage in cover system.
Storage of water in soil coupled with removal of water by ET are the most important
mechanisms for limiting percolation of infiltration. For most cover systems, infiltration is
primarily removed from the cover system by ET. Flow from lateral drainage layers is typically a
DRAFT - DO NOT CITE OR QUOTE
4-2
-------
much smaller component of the water balance than is ET. It should be remembered, however,
that while the internal drainage layer is typically of secondary importance to the overall cover
system water balance, it is of prime importance to cover system slope stability (see Chapter 6 of
this document). If even a relatively small amount of potential lateral flow is left undrained in a
cover system, hydraulic heads can build up over the hydraulic barrier, leading to destabilizing
seepage forces on cover system slopes.
Though Eq. 4.1 appears simple, the components of the water balance are dependent on many
factors, are difficult to quantify, and are interdependent. It can be especially difficult to quantify
percolation in arid and semi-arid environments where almost all precipitation is consumed by
ET. Unlike in wetter climates where actual ET may approach the magnitude of potential
evaotranspiration (PET) (i.e., the process is energy limited), in drier climates actual ET is
generally much smaller than PET due to the lack of available water. ET is more difficult to
accurately estimate under water limiting conditions. Because the magnitude of percolation in
drier climates is so much smaller than the magnitudes of ET and precipitation, relatively small
errors in estimated ET can result in relatively large errors in estimated percolation. Due to the
difficulty in performing accurate analytical water balances, field water balances have
occasionally been performed using cover system test plots to better assess the water balances
components (e.g., the ACAP program, as described in Section 3.4.3). For example, field water
balances have been performed for alternative cover systems without GM barriers and for cover
systems at low level radioactive waste containment and disposal sites. Examples where field
methods have been used to investigate one or more components of a cover system water balance
include Cartwright et al., 1988; Nyhan et al., 1990; Anderson et al, 1993; Gee et al., 1994;
Limbach et al, 1994; Melchior et al., 1994; Waugh et al., 1994; Dwyer, 1995; Khire, 1995;
Sackschewsky et al., 1995; Schultz et al, 1995; Paige et al., 1996; Anderson, 1997; Gee et al.,
1997; Karr et al., 1997; Khire et al., 1997; Laundre, 1997; Melchior, 1997a,b; Morris and
Stormont, 1997; Nyhan et al, 1997; Ward and Gee, 1997; Dwyer, 1998; Khire et al., 1999;
Dwyer, 2001; and Scanlon et al., 2002.
Water balance calculations are performed for time intervals that may be shorter than one hour or
longer than a year. The time interval to use is dependent on the purpose of the water balance
analysis. Guidance on the time interval to use for design is given subsequently.
4.2.3 Water Balance Methods
A variety of water balance methods are available to evaluate and design cover systems. They
range in complexity from relatively simple empirical correlations to sophisticated computer-
based finite difference and finite element mechanistic models. This guidance document
describes the following water balance analysis methods: (i) simplified manual method; (ii)
Hydrologic Evaluation of Landfill Performance (HELP) model; (iii) Leachate Estimation and
Chemistry Model (LEACHM); (iv) UNSAT-H model: (v) SoilCover model; and (vi) HYDRUS-
2D model. These are all well-documented manual methods or computer codes that consider the
significant water balance processes (e.g., precipitation, runoff, and ET) and that have been used
previously for cover system water balance analyses. All of the models except HYDRUS-2D are
in the public domain. The characteristics of these models are compared in Table 4.1.
DRAFT - DO NOT CITE OR QUOTE
-------
4.2.3.1 Simplified Manual Method
Koerner and Daniel (1997) present an updated version of the simplified method for performing
manual or computer spreadsheet water balance calculations for cover systems. Their method is
based on the previous work of Thornthwaite and Mather (1955, 1957), Fenn et al. (1975), and
Kmet (1982). In this previous work, only monthly time steps were considered. Historically,
simplified water balances using monthly time steps were used for cover system analysis and
design. The computer code MBALANCE (Scharch, 1985), based on the simplified manual
method with a monthly time step, was developed for landfill cover systems by Wisconsin
Department of Natural Resources. This model was used in simulations that were compared to
field water balances (Lane et al., 1992). Koerner and Daniel (1997) extended the method to
consider a variable time step (e.g., daily, weekly, or monthly) to be selected based on the
purpose of the analysis. A spreadsheet developed by Koerner and Daniel (1997) to evaluate
monthly percolation through a cover system is shown in Table 4-2. The table is readily
adaptable to PC-based spreadsheet computations and can be easily modified to accommodate
daily or hourly time steps. Guidance on, and an example of, the use of Table 4-2 are presented
in Koerner and Daniel (1997). The equation numbers given in the table are from that reference.
The remainder of this section addresses several important aspects of the simplified manual
method.
In the simplified manual method, it is assumed that no water is stored at the surface or
intercepted by plants (i.e., AWsurface = AWf0iiage = 0). For this set of assumptions, the following
relationships are defined for a time interval taken as one day:
P = I + R (Eq. 4.2)
I = ET + AWsoii + PERC* (Eq. 4.3)
where: I = infiltration into cover soil (mm/day); and PERC* = percolation through cover soil
(mm/day); and other terms are as defined previously.
In the simplified manual method, precipitation is partitioned into runoff and infiltration (Eq. 4.2).
Runoff is calculated as a fraction of precipitation using the rational formula and a runoff
coefficient appropriate for the cover system soil type and slope. According to Fenn et al. (1975),
the rational formula will, in most cases, underestimate the quantity of cover system runoff.
From Eq. 4.3, water infiltrating the cover soil is partitioned into ET, soil water storage, and
percolation through the cover soil. In the simplified manual method, ET is calculated as a
function of PET, infiltration, and initial moisture content of the soil. PET is calculated using an
empirical method developed by Thornthwaite and Mather (1955). If more water infiltrates the
cover system than can potentially evapotranspire, the excess water will first be distributed within
the root zone until the soil moisture content is at field capacity. The remaining water will be
routed as percolation through the cover soil. If ET is greater than infiltration, then stored water
will be lost from the cover soil root zone until the soil moisture content is at wilting point.
DRAFT - DO NOT CITE OR QUOTE 4-4
-------
Table 4-1. Comparison of select water balance models.
Model
Simplified
Manual
Method
HELP
LEACHM
UNSAT-H
Reference
Koerner
and Daniel
(1997)
Schroeder
etal. (1994a,
1994b)
for EPA
Hutson and
Wage net
(1992)
for Cornell
University
Payer and
Jones (1990)
for Pacific
Northwest
Laboratory
Calculation Scheme
Simplified empirical and
mechanistic equations
Quasi 2-D water-routing
model with multiple
uncoupled subroutines
Simplified empirical and
mechanistic equations
Simplified unsaturated
flow model with unit
hydraulic gradient
Finite difference model
with unsaturated flow
model based on
Richards' partial
differential equation
User specified boundary
conditions
Finite difference model
with unsaturated flow
model based on
Richards' partial
differential equation
User specified boundary
conditions
Advantages
Easy to perform
Few data requirements
Any time step
Considers lateral drainage
Widely accepted
Used to design hydraulic
barriers
Easy to run simulations
Default database of
climatic, soils, and
vegetation data
Considers lateral drainage
Mechanistic model
Solves unsaturated flow
equation
May give a better estimate
of ET in arid climates than
other models
Mechanistic model
Solves unsaturated flow
equation
Flexibility in definition of
unsaturated hydraulic
conductivity-head-
moisture content
relationships
Disadvantages
Numerous simplifying
assumptions must be
made
Steady-state conditions are
assumed
Essentially all calculations
are uncoupled
Cannot be used for
unsaturated flow
Does not solve unsaturated
flow equations
Demonstrated
overprediction of
percolation in many cases
Limited to daily climatic
data
Maximum soil profile depth
of 2 m
Does not consider lateral
drainage
High computational
demands
Unsuitable for parametric
evaluation
Does not consider lateral
drainage
Appropriate Use
Instructional tool for design
of hydraulic barriers
Check of computer
simulations
Parametric evaluations
Calculation of peak lateral
drainage from cover
system
Design of hydraulic barriers
Regulatory compliance
demonstrations
Parametric evaluations
Calculation of peak lateral
drainage from cover
system
Design of ET and capillary
barriers (no lateral flow)
Parametric evaluations
Unsaturated flow analysis
Performance assessment
of ET and capillary
barriers (no lateral flow)
Calibration with field data
prior to making long-term
predictions
Unsaturated flow analysis
DRAFT - DO NOT CITE OR QUOTE
4-5
-------
Table 4-1. Comparison of select water balance models (continued).
Model
Reference
Calculation Scheme
Advantages
Disadvantages
Appropriate Use
SoilCover
SoilCover
(2000)
Finite element model with
unsaturated flow model
based on Richards'
partial differential
equation
User specified boundary
conditions
Mechanistic model
Solves unsaturated flow
equation
Calculates actual
evaporation based on
matric suction at soil
surface
Easy to create input files
with spreadsheet user
interface
Limited boundary condition
options
High computational
demands
Maximum of 8 soil layers
Maximum of 100 nodes
Does not consider lateral
drainage
Requires temperature input
Performance assessment
of ET and capillary
barriers (no lateral flow)
Unsaturated flow analysis
Hydrus
2-D
Simunek et
al. (1999)
for U.S.
Salinity
Laboratory
Two-dimensional finite
element model with
unsaturated flow model
based on Richards'
partial differential
equation
User specified boundary
conditions
Mechanistic model
Solves unsaturated flow
equation
Flexibility in definition of
unsaturated hydraulic
conductivity-head-
moisture content
relationships
Considers lateral flow and
anisotropy
Inverse estimation of
hydraulic properties from
measured data
Considers spatial
heterogeneity
High computational
demands
Does not include vapor flow
Does not calculate PET
from climatic data
Not in public domain
Performance assessment
of ET and capillary
barriers with lateral flow
Calibration with field data
prior to making long-term
predictions
Unsaturated flow analysis
DRAFT - DO NOT CITE OR QUOTE
4-6
-------
Table 4-2. Example spreadsheet for simplified manual water balance method (Koerner and Daniel, 1997).
Row
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
P
Q
Parameter
Avg. Monthly Temp, °C
Monthly Heat Index (Hm)
Unadjusted Daily PET (UPET), mm
Possible Monthly Duration of Sunlight (N)
PET, mm
Precipitation (P), mm
Runoff Coefficient (C)
Runoff (R), mm
Infiltration (IN), mm
IN - PET, mm
Accumulated Water Loss (WL), mm
Water Stored (WS), mm
Change in Water Storage (CWS), mm
Actual ET (AET), mm
Percolation (PERC), mm
Check (CK), mm
Percolation Rate (FLUX), m/s
Reference
Input Data
Eq. 4.7
Eqs. 4.8 and 4.9
Table 4. 3 or 4.4
PET = UPET - N
Input Data
See Table 4.1
R = P-C
IN = P-R
WL = Z(IN - PET)<0
Section 4. 3. 1.12
Section 4. 3. 1.1 3
Eq.4.16
Eq.4.18
Eq.4.19
Eq. 4.20
^
re
c
re
February
.c
e
re
Q.
<
>
re
0)
c
3
->
>
3
->
+-
M
3
O)
3
<
September
October
November
December
Total
DRAFT - DO NOT CITE OR QUOTE
4-7
-------
If water does not flow laterally through an internal drainage layer, percolation through the
hydraulic barrier is equal to percolation through the cover soil (i.e., PERC* = PERC).
Conversely, if lateral flow occurs:
PERC* = PERC + L (Eq. 4.4)
where all terms are as defined previously. In the simplified manual method, Eq. 4.4 is solved
iteratively since both PERC and L are a function of hydraulic head.
Assuming steady-state conditions, the maximum flow in the internal drainage layer is calculated
as:
£L £ (PERC* -PERC)
- 8.64x10'
where: qm = maximum flow rate in drainage layer per unit width perpendicular to the direction of
flow (m3/s/m);A, = slope length (m); and other terms are as defined previously. The hydraulic
transmissivity of the drainage layer must be adequate to accommodate this flow. The flow
capacity of drainage layers was discussed in 2.4.2.3. Hydraulic design of drainage layers is
discussed subsequently in Sections 4.5 and 4.6.
Koerner and Daniel (1997) recommend that the hydraulic requirements of a cover system
drainage layer be evaluated based on a single storm event. They conservatively suggest that, for
design, the cover soil above the drainage layer be assumed to be saturated and that percolation
through the cover soil be set equal to infiltration into the cover soil (i.e., ET = 0 and AWsoii = 0).
For these conditions:
PERC* = P - R (Eq. 4.6)
where all terms are as defined previously. Applying the rational formula to the calculation of R
leads to:
PERC* = P (1 - Cr) (Eq. 4.7)
where: Cr = runoff coefficient (dimensionless) obtained from Table 4-3 or project-specific
information.
Eq. 4.7 was developed assuming that: (i) the cover soil is at field capacity before the storm
begins; (ii) there is no ET during the storm; and (iii) the cover soil is sufficiently permeable to
accept the calculated infiltration. To account for this last condition, Koerner and Daniel (1997)
suggest that PERC* calculated with Eq. 4.7 be adjusted in accordance with Thiel and Stewart
(1993) using Eq. 4.8a or 4.8b, depending on a comparison of the rate at which water becomes
available for infiltration to the saturated hydraulic conductivity of the cover soil.
DRAFT - DO NOT CITE OR QUOTE 4-8
-------
Table 4-3. Runoff coefficients (from Fenn et al., 1975) suggested by Koerner and Daniel
(1997) for simplified manual water balance calculations.
Soil Description
Sandy Soil
Sandy Soil
Sandy soil
Clayey Soil
Clayey Soil
Clayey Soil
Slope Runoff coefficient
Flat (< 2%)
Average (2 - 7%)
Steep (> 7%)
Flat (< 2%)
Average (2 - 7%)
Steep (> 7%)
0.05-
0.10-
0.15-
0.13-
0.18-
0.25-
0.10
0.15
0.20
0.17
0.22
0.35
PERC* = P(l - Cr) when kcs > P(l - Cr) (Eq. 4.8a)
PERC* = kcs when kcs > P(l - Cr) (Eq. 4.8b)
where: kcs = the cover soil saturated hydraulic conductivity in the same units as P. Eq. 4.8 can be
used to develop a conservative estimate of peak flow into a lateral drainage layer during a single
storm event, a capability available in only one (i.e., HYDRUS-2D) of the other water balance
models considered in this chapter.
In the simplified manual method, percolation through CCL or GCL barriers is calculated using
Darcy's equation, which describes the flow of fluids through porous media. Percolation through
GM and composite liners is calculated by Koerner and Daniel using the leakage rate equations
developed by Giroud and Bonaparte (1989a,b). Hydraulic head is an input parameter to these
equations. It is suggested that the maximum hydraulic head calculated on a monthly basis (hm as
derived subsequently) be used to calculate leakage rates through hydraulic barriers.
Input data needs for the simplified manual method are minimal. Only precipitation and mean
temperature data are required. Koerner and Daniel (1997) provide guidance for selecting all
other parameters (e.g., runoff coefficient, root zone depth, and soil water storage capacity). The
advantages of the method are its simplicity, ability to use a variable time step, and ability to
calculate lateral flows in cover system drainage layers. The main disadvantages of the method
are the steady-state nature of all calculations and the numerous simplifying assumptions.
Nonetheless when appropriately used, the simplified manual method presents an acceptable
approach to the design of hydraulic barrier type final cover systems. The method is in no way
adequate as a simulation or predictive tool, nor is it applicable to the analysis or design of
capillary barriers or ET barriers.
4.2.3.2 HELP
DRAFT - DO NOT CITE OR QUOTE 4-9
-------
The HELP computer code was developed by the U.S. Army Corps of Engineers Waterways
Experiment Station (WES) for EPA to enable design engineers to compare the relative hydraulic
performance of alternative waste containment system designs (Schroeder et al., 1994a, 1994b).
Increasingly, HELP is being used to calculate percolation rates through cover systems and peak
hydraulic heads in cover systems for slope stability analyses. HELP has been updated
extensively since its inception. At the time of this writing, HELP Version 3.07 is the most recent
revision. The documentation for HELP by Schroeder et al. (1994a, 1994b) can be purchased
from the National Technical Information Service ((800) 553-6847), downloaded from the
USEPA website at http://www.epa.gov/cincl/, or downloaded from the WES website at
http://www.wes.army.mil/el/elmodels. The most recent version of the code can be downloaded
from the WES website. Additional guidance on using HELP to evaluate landfill hydrologic
performance can be found in EPA (1991). Users should use the most current version of the
HELP model at the time the analysis is to be performed. Users should also recognize that
conclusions drawn from studies using older versions of the model may not be the same as the
conclusions that would be drawn using the most current version of the model.
The HELP model simulates hydrologic processes for landfills by performing sequential water
balance calculations using a quasi-2-D, gradually varying approach. According to Peyton and
Schroeder (1993), the model is considered quasi 2-D because it considers only vertical flow in
all layers except lateral drainage layers, where flow can be vertical or lateral. The model is
considered gradually varying because the simulation moves through time with the water balance
processes being considered steady over each time step. A conceptualization of the HELP model
is presented in Figure 4-2. The model can be used to separately evaluate each subprofile shown
in Figure 4-2, including the complete cover system profile.
The hydrologic processes considered in the model include precipitation, surface-water storage
(i.e., storage as snow), interception of precipitation by foliage, surface-water evaporation, runoff,
snow melt, infiltration, plant transpiration, soil water evaporation, soil water storage, vertical
flow (saturated and unsaturated) through non-barrier soil layers, vertical percolation (saturated)
through soil barriers, vertical percolation (saturated) through GM and GM/soil composite
barriers, and lateral drainage (saturated). Five main routines are used in the HELP model to
estimate runoff, ET, vertical drainage to barriers, vertical percolation through soil barriers, and
lateral or vertical flow (saturated) through lateral drainage layers. Several other routines interact
with the main routines to generate daily precipitation, temperature, and solar radiation values and
to simulate snow accumulation and melt, vegetative growth, interception, and vertical
percolation through GM and GM/soil composite barriers.
Runoff in the HELP model is computed using the runoff curve number method of the USDA
SCS) (SCS, 1985). (Note that the Soil Conservation Service (SCS) is now the Natural Resources
Conservation Service (NRCS).) The method empirically correlates total runoff with total rainfall
based on daily rainfall records, vegetation type, soil type, antecedent moisture conditions (level
of soil moisture prior to rainfall), and other factors. The method does not consider the time
DRAFT - DO NOT CITE OR QUOTE 4-10
-------
Precipitation
Runoff Evapotranspi ration
r,%
." Q.
Cfl
0
g 0
» E"
m en
0
•F o
I--S
W
Vertical
Percolation Layer
Lateral Drainage Layer
SM. Earlier •
CCL Barrier
Vertical
Percolation Layer
Lateral Drainage Layer
GM Barrier '
Lateral Drainage Layer
Geomembrane Barrier *
CCL Barrier
— Percolation
—^^ Waste
."'.' .".'•''•'•' ' Sand '••'.'•.'• ' '•
. • ' Lateral Drainage '•'.'•-'.•.'..•
(Leachate Collection)'_•'' ''
• : ••'-.- ^*~^:' • i' I, ' •
Late'ral Drainage V • '. .' : Leakage '.
_ (Leakage Detection) 'sand . •'• .'• • ' '. '. '.
Leakage
Waste
Liner
System
Figure 4-2. Conceptualization of HELP water balance model (from Schroeder
etal., 1994a).
distribution of rainfall intensity and, therefore, does not give accurate estimates of runoff
volumes for individual storm events. The daily runoff is calculated in the model as:
R
(P-0.2Sr)2
(P + 0.8Sr)
(Eq. 4.9)
where: Sr = retention parameter (mm/day) dependent on SCS curve number; and R and P are as
defined previously. The SCS curve number is a function of soil texture, vegetation quality, and
cover system slope length and inclination. Schroeder et al. (1994a) indicate that long-term
cumulative runoff should be independent of rainfall duration and intensity, since over a long
simulation period a variety of precipitation events will occur. However, McBean et al. (1995)
state that use of daily rainfall averages effectively decreases storm intensity (because the
duration of most storms is less than 24 hours), resulting in a simulation having an overprediction
of infiltration and underprediction of runoff.
ET is computed in HELP by a two-stage modified Penman energy balance method developed by
Ritchie (1972). This method uses the PET concept as the basis for prediction of surface and soil
water evaporation and plant transpiration. The PET demand is first met by evaporation of water
or snow on foliage or on the ground, then soil water evaporation, and finally plant transpiration.
ET is assumed to occur within the evaporative zone depth specified by the user and is not
allowed to occur within or below a barrier. Also, the soil water content is not allowed to
DRAFT - DO NOT CITE OR QUOTE
4-11
-------
decrease below the wilting point, which is defined in the model as the volumetric water content
at a matric potential of -1.5 MPa. Due to these controls, ET may be underestimated in arid
climates. Growth and decay of surface vegetation is modeled using an algorithm taken from the
Simulator for Water Resources in Rural Basins (SWRRB) model (Arnold et al., 1989).
Vertical drainage for cover soil (i.e., topsoil and protection) layers for both saturated and
unsaturated flow conditions is computed using Darcy's equation. HELP assumes that soil
pressure head is constant within a vertical percolation layer. Changes in either positive or
negative pressure head cannot be modeled. The hydraulic gradient is due to change in elevation
head only and is thus equal to 1.0. The HELP model does, however, define an unsaturated
hydraulic conductivity to use with the unit hydraulic gradient for calculating unsaturated flow
rates. The unsaturated hydraulic conductivity, ku (m/s), is obtained in the HELP model using
Campbell's equation (1974):
ku=ks[(0-0r)/(0s-0r)]3+2/x (Eq.4.10)
where: ks = saturated hydraulic conductivity of soil layer (m/s); 0 = volumetric water content of
soil layer (dimensionless); 0S = volumetric water content of soil layer at saturation
(dimensionless); 0r = residual volumetric water content, typically in the range of 0.01 to 0.10
(dimensionless); and X = pore size distribution index (dimensionless), calculated as described in
Schroeder et al. (1994a,b). As a result of the hybrid formulation given above, the HELP model
cannot be used to simulate the physics of water movement through an unsaturated soil layer.
Lateral drainage below a cover soil layer is modeled by an analytical approximation to the
steady-state solution of the Boussinesq equation. The peak daily head in a drainage layer is
calculated using an equation formulated by McEnroe (1993). Vertical percolation through low-
permeability soil hydraulic barriers is evaluated in HELP using Darcy's equation assuming
saturated conditions. Vertical percolation through GMs and GM/soil composite barriers is
evaluated based on the work of Giroud and Bonaparte (1989a,b) and Giroud et al. (1992a).
The daily water balance is calculated in the HELP model by a linking process, starting with a
surface water balance, then ET in the subsurface, and finally subsurface water routing from the
surface downward one soil layer at a time. The routing procedure uses a time step that can range
from 30 minutes to six hours. However, only daily, monthly, annual, and long-term average
output data are reported.
The HELP model requires daily and general climatic data, material properties data for the
landfill components being modeled, and landfill design data. One of the strengths of the HELP
model is its climatic and material property default data option. Required daily weather data are
precipitation, mean temperature, and total global solar radiation. Daily precipitation may be
input manually, selected from a historical database (e.g., 1974-1977 data in the HELP database,
NOAA Tape, or Climatedata™ files), or generated stochastically using a weather generation
model developed by the U.S. Department of Agriculture-Agricultural Research Service (USDA-
ARS) (Richardson and Wright, 1984) with simulation parameters available for 139 U.S. cities. It
should be noted that the historic precipitation data in the database for 1974-1977 are often not
DRAFT - DO NOT CITE OR QUOTE 4-12
-------
used because they are for an unusually dry time period in certain parts of the U.S. Other daily
climatologic data are generated stochastically using the USDA-ARS routine. Required general
weather data include average annual wind speed and latitude. Default general weather data for
183 U.S. cities are used by the model. The material properties of each layer being modeled are
either selected from the HELP model database of default material properties or are specified by
the model user. Landfill design data, including landfill general information and layer
configuration, are user specified.
Due to its method of calculating downward flux and its limiting of upward flux (i.e., no upward
flux within or below a barrier), version 3.07 of the HELP model is not considered a particularly
accurate simulation model for cover systems located in arid areas where the subtleties of
unsaturated moisture movement can dominate the water balance. As will be discussed, there are
other water balance models that better simulate the physics of water movement in arid
environments.
4.2.3.3 LEACHM Model
LEACHM (Hutson and Wagenet, 1992) is a one-dimensional finite difference code that is
finding increasing use in the western United States, particularly California, for design and
performance analysis of cover systems with ET barriers. LEACHM was originally developed to
simulate the effects of agricultural management alternatives on the movement of water and
chemicals in a shallow soil profile (i.e., to a maximum depth of 2 m). Only the hydrologic
component of the model will be discussed further. The code and model documentation may be
obtained from the Department of Soil, Crop & Atmospheric Sciences at Cornell University,
Ithaca, New York.
The LEACHM model considers precipitation, runoff, ET, soil water storage, and percolation in
the water balance. Infiltration of water into the soil profile and vertical drainage are simulated
using a finite difference solution to Richards' partial differential equation (Richards, 1931). This
equation is obtained by combining the differential form of Darcy's equation for unsteady vertical
flow with the one-dimensional differential form of the conservation of mass equation:
dt dz
-S(z,t) (Eq. 4.11)
where: \j/ = matric potential (negative) due to capillary suction forces (N/m2); 9 = soil volumetric
water content (dimensionless); ku = unsaturated hydraulic conductivity (m/s); z = vertical
coordinate, positive downward (m); t = time (s); and S(z,t) = sink term representing uptake by
transpiration (s"1).
Unsaturated soil hydraulic conductivity in LEACHM is calculated using the Campbell (1974)
relationship. Precipitation in excess of the infiltration capacity of the soil is shed as runoff.
Evaporation and transpiration are modeled separately based on the methods of Childs and Hanks
(1975). With this method, the potential evaporation and transpiration are first estimated based
on the pan evaporation rate, a pan factor, and a crop cover fraction. The actual evaporation is
then calculated as the lesser of the potential evaporation and the possible evaporation calculated
DRAFT - DO NOT CITE OR QUOTE 4-13
-------
using Richards' equation and the selected boundary condition. Any remaining PET demand is
applied to transpiration. However, transpiration is not allowed if the matric potential head of the
soil is less than -1.5 MPa, the potential assumed to correspond approximately to the soil wilting
point.
LEACHM requires that climatic data, soil properties, vegetation data, and initial and boundary
conditions be input. Unlike the HELP model, there are no default data; the user must specify
each input parameter. Required weather data are precipitation magnitude, rate, and start time,
minimum and maximum air temperatures, and pan evaporation rate. The precipitation option
allows rainfall data for short, intense storms to be input. Thus, LEACHM may be used to
estimate the head of water in the cover system due to a design storm. In the absence of pan
evaporation rate data, the rate can be calculated by LEACHM using the Linacre equation
(Hutson and Wagenet, 1992) with site-specific data (i.e., latitude, elevation, temperature, and
precipitation). Required soil data are bulk dry density, initial moisture content, saturated
hydraulic conductivity, and soil water retention curve. If a soil water retention curve is not
available, LEACHM contains a routine to compute fitting parameters for Campbell's soil-water
retention curve from the particle size distribution, bulk density, and organic matter content of the
soil. However, there is considerable uncertainty in the use of the regression equations to
compute these parameters. Vegetation data to be input are root depth and distribution, plant
growth options (i.e., constant vegetation or growing vegetation), wilting point, minimum root
potential, maximum ratio of actual to potential transpiration, root resistance, and plant growth
timeline (e.g., germination, emergence, maturity, etc.). Very little guidance is provided in the
LEACHM model documentation on selection of values for the various input parameters.
To set up the finite difference grid used by LEACHM, the soil profile is divided into a number of
horizontal layers of equal thickness with nodes at the center of each layer. Soil properties are
then specified for each layer. Two additional nodes are required for boundary conditions, one
above the ground surface and one below the profile being modeled. The upper boundary
condition can be changed with time by adjusting the head to simulate ponded or non-ponded
infiltration, evaporation, or zero flux. The lower boundary condition can be selected as a fixed
water table, free drainage (or unit gradient), zero flux, or lysimeter boundary. The initial
condition is specified by assigning an initial head or water content to each node in the finite-
difference nodal grid. Simulation output includes cumulative infiltration, evaporation,
transpiration, and percolation at select times.
4.2.3.4 UNSAT-H
UNSAT-H is a one-dimensional finite-difference water balance model developed at Pacific
Northwest Laboratory (Payer and Jones, 1990) to assess the water dynamics of waste disposal
facilities at the U.S. Department of Energy (DOE) Hanford site. The model also simulates soil
heat flow and nonisothermal vapor flow. Vapor flow can be an important transport mechanism
in near surface soils at arid sites. The UNSAT-H model was derived from the UNSAT model of
Gupta et al. (1978) and has retained many of the same routines. At the time of this writing,
Version 3.0 of UNSAT-H was the most current. The code can be obtained from the Energy
Science and Technology Software Center, Department of Energy, Oak Ridge, Tennessee.
DRAFT - DO NOT CITE OR QUOTE 4-14
-------
The UNSAT-H model considers precipitation, runoff, ET, soil water storage, and percolation in
the water balance. Like the LEACHM model, infiltration of water into, and vertical movement
of moisture in, the soil profile is governed in the UNSAT-H model by a finite difference solution
to Richards' partial differential equation. However, the unsaturated soil hydraulic conductivity
term in the UNSAT-H model is calculated using polynomials, Haverkamp functions, Brooks-
Corey functions, or van Genuchten functions rather than the Campbell equation. Precipitation in
excess of the infiltration capacity of the soil is shed as runoff. Evaporation and transpiration are
considered separately.
Evaporation in the UNSAT-H model is calculated using one of two approaches: (i) an integrated
form of Pick's law of diffusion that considers the flow of heat to and from the soil surface, the
flow of water from the subsurface to the soil surface, and the transfer of water vapor from the
soil surface to the atmosphere; or (ii) a Penman-type equation that is a modification of the
diffusion equation and is dependent on net radiation and soil heat flux rather than on soil-surface
temperature. Transpiration is calculated using a method based on leaf-area index or cheatgrass
data and is limited by PET.
The UNSAT-H model requires that climatic data, soil properties, vegetation data, and initial and
boundary conditions be input. There are no default data; the user must specify each input
parameter. Required data for the meteorological data option are daily precipitation, daily
maximum and minimum air temperatures, daily solar radiation, average daily dew point, and
average daily wind speed. Daily precipitation and PET may be input instead of daily
meteorological data. The precipitation option allows rainfall data for short, intense storms to be
input. Required soil data are fitting parameters for the soil water characteristic functions and the
unsaturated hydraulic conductivity functions. An option for including hysteresis is available.
Vegetation data to be input include root depth, leaf area index, growing season, and percent bare
area. Very little guidance is provided in the UNSAT-H model documentation on selection of
values for the various input parameters.
The finite difference grid used by UNSAT-H is set up in a manner similar to that for LEACHM.
The soil profile is divided into a number of horizontal layers with nodes located at the center of
each layer. Two additional nodes, one above the ground surface and one below the profile being
modeled, are used to set boundary conditions. The upper boundary condition can be changed
with time by adjusting the head to simulate ponded or non-ponded infiltration, evaporation, or
zero flux. The lower boundary condition can be selected as a fixed water table, free drainage (or
unit gradient), zero flux, or specified flux boundary. The initial condition is specified by
assigning an initial head or water content to each node in the finite-difference nodal grid.
Simulation output includes infiltration, evaporation, transpiration, and percolation at hourly or
daily intervals.
4.2.3.5 SoilCover
SoilCover model was developed in 1990 at the University of Saskatchewan for the analysis of
the flow of water and heat between the atmosphere and the soil surface, particularly for land
based disposal systems. Since then the model has been modified by Geo-Analysis 2000 Ltd.,
Saskatoon, Canada to include oxygen diffusion, an enhanced vegetation routine, freeze/thaw
DRAFT - DO NOT CITE OR QUOTE 4-15
-------
considerations, and soil property function revisions. SoilCover Version 5.2 was the most recent
release at the time of this writing. The code and accompanying user's manual is available for
download from http://www.members.shaw.ca/geo2000/pagel2.html.
SoilCover uses a finite-element method to solve the one-dimensional heat and mass transfer
partial differential equations derived by Wilson (1990). The mass transfer equation is obtained
by combining the differential forms of Darcy's law and Pick's law for unsteady vertical flow
with the one-dimensional differential form of the conservation of mass equation. Both liquid
flow and nonisothermal vapor flow are incorporated into the model. There is no option for
isothermal vapor flow, nor is there an option for shutting off vapor flow altogether like is
available with UNSAT-H. The unsaturated hydraulic conductivity function in the SoilCover
model may be either user-defined (i.e., tabulated data) or predicted based on a Fredlund-Xing
curve (Fredlund and Xing, 1994) fit to the water content versus matric potential data. The
method used to predict the unsaturated hydraulic conductivity function was developed by
Fredlund et al. (1994), and, according to SoilCover (2000), is especially well-suited for modeling
fine-grained soils. Precipitation, runoff, ET, soil water storage, and percolation are included in
the water balance.
SoilCover calculates evaporation using a modified Penman equation developed by Wilson
(1990):
FRn+v
E =
(Eq. 4.12)
where: EV = vertical evaporative flux (mm/day); F = slope of the saturation vapor pressure versus
temperature curve at the mean temperature of the air (dimensionless); Rn= net radiant energy
available at the surface (mm/day); v = psychrometric constant (dimensionless); Ua= wind speed
(km/hr); Pa = vapor pressure in the air above the evaporating surface (Pa); ha = relative humidity
of the air (dimensionless); and hr = relative humidity at the soil surface (dimensionless). The
model also offers the option of calculating evaporation based on user-input PET, in which case it
uses the following equation:
E,, = PET
(Eq. 4.13)
where all terms are as defined previously. Unlike the other models described in this report,
SoilCover calculates evaporation as a sink term directly from the surface relative humidity,
which is a function of the matric suction and the temperature at the soil surface. The developers
of SoilCover claim this method of calculating evaporation is a strength of the model.
Runoff is calculated as any precipitation that cannot infiltrate. Transpiration is calculated by
DRAFT - DO NOT CITE OR QUOTE 4-16
-------
applying fluxes at nodes in the root zone. Plant water stress and canopy shading effects are also
considered by SoilCover.
The SoilCover model requires that climatic data, soil properties, vegetation data, and initial and
boundary conditions be input Required climatic data include daily maximum and minimum air
temperature, daily net radiation, daily maximum and minimum relative humidity, and daily wind
speed. If the option for entering daily PET is chosen, then daily net radiation and wind speed are
not required. Precipitation is entered on a daily basis as a constant flux top boundary condition,
but intensity may be accounted for by constraining the precipitation between specified hours.
Climatic data input is relatively easy because of the SoilCover's Microsoft Excel user interface.
Daily data may be copied from a spreadsheet source and pasted directly into SoilCover.
Properties for up to eight soils may be entered. Required soil properties are porosity, specific
gravity, saturated hydraulic conductivity, and coefficient of volume change. In addition, up to
20 water content vs. suction data points may be input. SoilCover then fits the Fredlund-Xing
(1994) soil-water characteristic function to the data points. The unsaturated hydraulic
conductivity function, the thermal conductivity function, and the volumetric specific heat
function can then be generated using the fit soil-water characteristic function. The user may also
choose to enter tabulated data for these functions. Very little guidance is provided in the
SoilCover user's manual on selection of values for the various input parameters, however a short
list of coefficients of volume change for typical soils is provided. Required input parameters for
vegetation include growing season start and stop day, moisture wilting and limiting points, daily
depths to top and bottom of roots, and selection of either poor, good, or excellent grass quality
The bottom boundary condition may be specified as either constant pressure or constant water
content. There is no option for constant flux, constant gradient, or seepage face lower boundary
conditions. The sparse lower boundary condition options necessitate that the user pay very close
attention to the bottom boundary fluxes throughout the duration of the simulation to ensure that a
realistic boundary is being modeled. For many landfill cover simulations, including a coarse-
grained soil beneath the soil profile and adjusting the value of the bottom boundary condition is
necessary to avoid "wicking" water from the boundary condition itself. If a gravel layer is added
below the profile, percolation results may be obtained by utilizing the SoilCover option of
cumulating fluxes at a selected internal node. The bottom temperature boundary condition must
also be specified on a daily basis.
The finite element mesh is generated by SoilCover from input depths and thickness of the soil
layers. Maximum and minimum node spacing for each layer must be specified along with the
node spacing expansion factor. Only 100 nodes are permitted, so spacing and expansion factors
may need to be adjusted. Initial conditions (either water content or suction) are also assigned to
each node based on the initial conditions input for the top and bottom of each layer. SoilCover
linearly interpolates the initial conditions. However, the assigned initial conditions may be
overwritten by the user after the mesh has been generated. Simulation output includes
infiltration, evaporation, transpiration, and percolation at daily intervals.
DRAFT - DO NOT CITE OR QUOTE 4-17
-------
4.2.3.6 HYDRUS-2D
HYDRUS-2D is a two-dimensional unsaturated flow model developed at the U.S. Salinity
Laboratory (Simunek et al., 1999). The model also simulates heat flow and solute transport.
The current model is an extension of the earlier unsaturated flow codes SWMS_2D and
CHAIN_2D. At the time of this writing version 2.02 of HYDRUS-2D was the most current. The
model may be purchased from the International Ground Water Modeling Center, Colorado
School of Mines, Golden, Colorado or
http://www.Mines.EDU/research/igwmc/software/igwmcsoft/. The documentation and a free
demo version of HYDRUS-2D may be downloaded from
http://www.ussl.ars.usda.gov/models/hydrus2d/htm.
HYDRUS-2D uses a finite element method to solve Richards' equation in a plane oriented either
vertically or horizontally. The two-dimensional domain may take on any geometric shape.
Because the model is two-dimensional, lateral flow and anisotropy may be simulated. A sink
term is included in Richards' equation for removal of water via plant transpiration. Vapor flow
cannot be simulated. The model has an option for allowing soil properties to be temperature
dependent, and it also allows hysteresis and spatial variability through a scaling transformation.
The unsaturated hydraulic conductivity is calculated by either a Brooks-Corey, van Genuchten-
Mualem, or modified van Genuchten method. Precipitation, runoff, ET, soil water storage, and
percolation are included in the water balance.
Precipitation and potential evaporation are the only climatic inputs required. HYDRUS-2D does
not have an option for internally calculating potential evaporation, so the user must use another
model or method to generate data to input. Vegetation parameters required include the heads
between which transpiration occurs and also the heads between which transpiration is optimal.
A menu containing a variety of properties for plants is available. The distribution of roots must
also be specified. Input required for soil properties includes saturated hydraulic conductivity and
fitting parameters from the selected soil-water retention function. A menu of soil properties is
available. In addition, van Genuchten properties can be predicted by inputting the percentage of
sand, silt and clay, density, field capacity, and/or wilting point water content. HYDRUS-2D also
has the option for inverse estimation of soil hydraulic properties from measured flow data.
The two-dimensional profile is created through a pre-processing module called Meshgen2D
within the HYDRUS-2D graphical user interface. After the domain geometry is defined,
Meshgen2D assists in generating the finite element mesh.
Boundary conditions may be specified flux, specified pressure head, unit gradient, atmospheric,
seepage face, or deep drainage. Precipitation and potential evaporation are specified using the
atmospheric option, which allows the boundary condition at the soil surface to change from
either prescribed flux or prescribed head. The user inputs the upper and lower limits of head for
which the prescribed flux boundary operates. Therefore, evaporation and precipitation will
proceed at the potential rate until the soil surface dries or wets to a specified head. Once below
the specified head, the boundary changes to a prescribed head boundary condition, and
evaporation is limited by the ability of water to flow to the surface. If the surface becomes
saturated during precipitation, excess precipitation is removed as runoff. The seepage face
DRAFT - DO NOT CITE OR QUOTE 4-18
-------
option allows water to exit the domain when the soil adjacent to the boundary becomes saturated.
Deep drainage provides an option for a variable flux depending on the level of the groundwater
table. Initial conditions may be specified as either water contents or pressure heads.
The HYDRUS-2D post-processor allows a variety of options for viewing output. Results can be
displayed graphically, including an animation of changes in pressure head or water content
through time. Cross-sections plotting pressure head or water content vs. depth or length may be
taken from the profile at any time of the simulation. Other output options include viewing the
instantaneous or cumulative water boundary fluxes over time, run time information, graphical
display of soil hydraulic properties, or converting output to ASCII format.
4.3 Evaluation of Water Balance Models
4.3.1 Overview
A number of researchers have performed field studies and analytical assessments to evaluate the
HELP, LEACHM,UNSAT-H, SoilCover, and HYDRUS-2D models (Thompson and Tyler,
1984; Peters et al., 1986; Barnes and Rodgers, 1988; Peyton and Schroeder, 1988; Nyhan, 1989;
Wilson, 1990; Nichols, 1991; Udoh, 1991; Payer et al., 1992; Lane et al., 1992; Benson et al.,
1993; Peyton and Schroeder, 1993; Martian, 1994; Tratch, 1994; Fleenor and King, 1995; Khire,
1995; Khire et al., 1997; Webb et al., 1997; Zornberg and Caldwell, 1998; Scanlon et al., 2002).
These studies were used to either simulate field or laboratory water balance data or to investigate
trends and magnitudes of the different water balance components (i.e., infiltration, runoff, etc.).
The conclusions of these studies are not always in general agreement. For example, some
studies found that a certain model overpredicted or underpredicted infiltration or percolation in a
certain climate, whereas, other studies using the same model concluded just the opposite. In
many of the comparisons between measured and calculated water balances, site-specific field
data were used in the water balance predictions. However, in the current state of practice for the
majority of projects, measurement of site-specific parameters required for the models, such as
soil field capacity, wilting point, and evaporation depth or rooting depth, is not performed. Thus,
the model user is left to depend on default data, which may lead to an inaccurate representation
of a site. At present, these hydrologic models should be used carefully to ensure a conservative
and reasonable basis for design. As a true predictive tool, the value of the models is limited
unless site-specific calibrations are performed. The results of a few of the more significant field
studies are presented below.
4.3.2 Lysimeters at DOE Hanford Site
Payer et al. (1992) compared field water balances for eight unvegetated lysimeters at DOE's
Hanford site to water balances simulated using the UNSAT-H, Version 2 model. The Hanford
site is located about 35 km northwest of Richland, Washington, in the northern cold desert of the
Columbia Basin. Average annual rainfall at the site is only 162 mm and average potential
evaporation is 1,600 mm (Gee et al., 1994). On average, over 70% of precipitation falls during
October through April. The soil profile in the lysimeters and the simplified profile used for
simulations are shown in Figure 4-3. The uppermost soil in the lysimeters is a silt loam material.
The soil profile in the lysimeters is intended to simulate a capillary barrier.
DRAFT - DO NOT CITE OR QUOTE 4-19
-------
Lysimeter Design
Conceptual Model
(UNSAT-H Code)
1.98m
3.05m
0.05m
0.05 m
O.OSTn
Silt Loam
Sand (20-30)
Sand (No. 8)
Gravel (10mm)
^Gravel (20 mm) f
, Railroad Ballast (40-50 mm)
1.50m
0.10m
Silt Loam
Sand
Gravel
(10mm)
Drain
Figure 4-3. Lysimeter design and conceptual model used to compare measured and
simulated water balance for DOE Hanford site (from Payer et al., 1992).
Of the eight lysimeters constructed by Payer et al. (1992), six were drainage lysimeters and two
were weighing lysimeters. The drainage lysimeters comprised two replicates of three
precipitation treatments: (i) ambient; (ii) two times the average annual precipitation; and (iii)
breakthrough (i.e., water added until drainage occurred). The weighing lysimeters served as
additional replicates, with one of the lysimeters receiving the normal precipitation and the other
receiving two times the average annual precipitation. Soil water content and percolation data
were collected for the lysimeters from November 1987 to April 1989.
The field water balances for the lysimeters were compared to water balance simulations
performed using IMSAT-H. The simulations were performed with actual weather data from a
nearby meteorological station, measured soil properties data for the silt loam, and assumed
properties for the sand and gravel layers beneath the silt loam. The lower boundary of the
drainage lysimeters was modeled as a unit gradient and the lower boundary of the weighing
DRAFT - DO NOT CITE OR QUOTE
4-20
-------
lysimeters was represented as a zero-flux condition. The upper boundary condition was allowed
to vary depending on climatic conditions.
Measured and simulated water contents for the drainage lysimeters under the three precipitation
conditions are shown in Figures 4-4 to 4-6. Measurable percolation only drained from the
lysimeters with the "breakthrough" precipitation treatment. In general, the simulated soil water
profiles showed reasonable agreement with measured water contents. However, UNSAT-H
tended to underestimate somewhat the amount of soil water storage during the spring and
overestimate the amount of soil water storage during the winter. Payer et al. (1992) attributed
this discrepancy primarily to the underestimation of evaporation in the winter and the
overestimation of evaporation in the summer. This effect is also apparent in the plot of measured
and simulated soil water storage in Figure 4-7(a). By decreasing evaporation, increasing the
saturated hydraulic conductivity of the silt loam, and adding a snow cover, simulated soil water
storage shows better agreement with measured soil water storage (Figure 4-7 (b)).
50
E
H 100
o
1
CO
0>
Q
150
200
250
300
(a)
2 Nov. 1988
Sim. Day 364
Measured Simulated
o
D
50
100
150
200
250
300
(b)
14 March 1989
Sim. Day 496
I
0.1 0.2 0.3 0.4 0 0.1
Water Content (cm3/cm3)
0.2
0.3
0.4
Figure 4-4. Measured and simulated (UNSAT-H) water contents for the ambient
precipitation treatment at DOE Hanford lysimeters (from Payer et al., 1992).
DRAFT - DO NOT CITE OR QUOTE
4-21
-------
50
100
w 15°
1
0
GO
.c
•s.
0
Q
200
250
300
(a)
2 Nov. 1988
Sim. Day 364
Lysimeter Measured
D10 •
D12 o
W4 D
Simulated
0.1
0.2
0.3
50-
100-
150
200.
250.
300
0.4 0
1 1
14 Mar. 1988
"Sim. Day 496
m
m
en •
JO] •
I
0.1
0.2
0.3
0.4
Water Content (cm3/cm3)
Figure 4-5. Measured and simulated (UNSAT-H) water contents for the 2x average
precipitation treatment at DOE Hanford lysimeters (from Payer et al., 1992).
50 -
3- 100-
•i
^
0)
m
o.
Q
200.
250 •
300
\
o
o
o •
0
c
•
1
^•^^
1 1
1 (a)
18 May 1988
Sim. Day 196
D
0*
L °*
I
50
100
150
200
250
300
29 June 1988
Sim. Day 238
Lvsimeter Measured
D9 •
Simulated
Desorp. Sorption
Curve Curve
0.1
0.2
0.3
0.4
0.1
0.2
0.3
0.4
0.5
Water Content (cm3/cm3)
Figure 4-6. Measured and simulated (UNSAT-H) water contents for the breakthrough
precipitation treatment at DOE Hanford lysimeters (from Payer et al., 1992).
DRAFT - DO NOT CITE OR QUOTE
4-22
-------
0)
o
I
V)
Q
&
£
'S.
&
50
100
150
200
250
300
0
\
0
o
0 •
c
c
_
I
I
I
I (a)
18 May 1988
Sim. Day 196
0
o«
L o*
I I
I
0.1 0.2 0.3 0.4
50
100
150
200
250
300
(b)
29 June 1988
Sim. Day 238
Desorp. Sorption
Curve Curve
D9
D11
0.1
Water Content (cm3/cm3)
0.2
0.3
0.4
0.5
Figure 4-7. Measured and simulated (UNSAT-H) water storage for the 2x average
precipitation treatment at DOE Hanford lysimeters: (a) initial simulation;
(b) simulation with improved calibration (from Payer et al., 1992).
4.3.3 Test Plots at Hill Air Force Base
Paige et al. (1996) described calibrating Version 2 of the HELP model to field measurements
from two cover system test plots constructed at Hill Air Force Base (Hill AFB), in Layton, Utah
and monitored for a four-year period. The calibrated models were then used to simulate the
long-term performance of the cover systems. One test plot had an ET-type soil cover ("control
soil cover") consisting of a 0.9-m thick sandy loam topsoil layer. The other test plot had a cover
system consisting of the following components, from top to bottom: 1.2-m thick sandy loam
topsoil layer; 0.3-m thick sand lateral drainage layer; and 0.6-m thick CCL. Both cover systems
were constructed over a 0.3-m thick gravel layer with lysimeters so that percolation could be
monitored. Cross sections of the cover systems are shown in Figure 4-8. After construction, the
plots were vegetated with native grasses. Water balance data measured over the four-year
monitoring period include precipitation, lateral flow in the sand drainage layer, percolation, soil
moisture content, and runoff.
Using the HELP model default values for the ET-type cover, HELP overpredicted annual ET by
approximately 30% and underpredicted annual percolation by approximately 95%. For the
hydraulic barrier-type cover, ET was overpredicted by 48%, runoff was overpredicted by 150%,
and lateral drainage was underpredicted by 97% when the HELP model was run with default
values. The HELP model was subsequently calibrated to the field water balances primarily by
modifying the soil properties of the cover systems (e.g., saturated hydraulic conductivity, soil
water storage capacity). The measured and calibrated values of the water balances for the ET-
type cover system and the hydraulic barrier-type cover system are shown in Tables 4-4 and 4-5,
respectively. As can be seen from these tables, even with the site-specific calibration, significant
DRAFT - DO NOT CITE OR QUOTE
4-23
-------
differences between field and simulated water balance components occurred. In particular, for
the ET cover system, correlation between measured and predicted percolation was not good.
(a) Control Soil Cover
-Neutron Probe Access Tube (typ)
0.9m
Runoff
Sediment
,_, Collection
Pan Lysimeter (typ)
(b) RCRA Cover
/— Neutron Probe Access Tube (typ)
Runoff
Sediment
Collection
1.2m
0.3 m
0.6 m
Lateral
Flow
Collection
Pan Lysimeter (typ)
Figure 4-8. Hill Air Force Base test plots: (a) ET-type cover system; and (b) hydraulic
barrier-type soil cover system (from Paige et al., 1996).
DRAFT - DO NOT CITE OR QUOTE
4-24
-------
Table 4-4. Difference between measured annual values and HELP simulation values for the control ET-type cover system at
Hill AFB (modified from Paige et al., 1996). Results obtained using input parameters calibrated from site water
balance data.
Water Balance
Variable
1991
Precip.
Runoff
Perc.
ET
Soil water1
1992
Precip.
Runoff
Perc.
ET
Soil water
1993
Precip.
Runoff
Perc.
ET
Soil water
Measured
(cm)
53.72
1.50
9.09
34.70
8.43
39.09
0.10
5.79
33.30
-0.10
41.78
0.25
23.80
30.66
12.93
(% meas. precip.)
100.00
2.79
16.93
64.58
15.70
100.00
0.26
14.81
85.18
-0.26
100.00
0.61
56.96
73.37
-30.94
HELP Predicted
(cm)
53.70
1.14
17.09
34.64
0.81
39.26
0.25
10.84
28.47
-0.30
41.85
0.61
10.49
32.18
-1.44
(% pred. precip.)
100.00
2.14
31.84
64.53
1.49
100.00
0.63
27.62
72.50
-0.75
100.00
1.49
25.08
76.88
-3.44
Difference
(cm)
-
0.36
-8.00
0.06
7.62
.
-0.15
-5.05
4.83
0.20
.
-0.36
13.31
-1.52
14.37
(% meas. precip.)
-
0.67
-14.90
0.11
14.19
.
-0.38
-12.92
12.36
0.51
.
-0.86
31.86
-3.64
34.39
Change in soil water storage.
DRAFT - DO NOT CITE OR QUOTE
4-25
-------
Table 4-5. Difference between measured annual values and HELP simulation values for the control soil cover system at Hill
AFB (modified from Paige et al., 1996). Results obtained using input parameters calibrated from site water
balance data.
Water Balance
Variable
1991
Precip.
Runoff
Lat. Drain
Perc.
ET
Soil water1
1992
Precip.
Runoff
Lat. Drain
Perc.
ET
Soil water
1993
Precip.
Runoff
Lat. Drain
Perc.
ET
Soil water
Measured
(cm)
53.72
1.14
19.00
0.00
24.59
8.99
39.09
0.05
6.70
0.00
30.12
2.21
41.78
0.71
23.32
0.00
27.94
-10.18
(% meas. precip.)
100.00
2.13
35.37
0.00
45.77
16.73
100.00
0.13
17.15
0.00
77.06
5.65
100.00
1.70
55.80
0.00
66.87
-24.37
HELP Predicted
(cm)
53.70
0.84
17.25
0.28
34.36
0.99
39.26
0.13
11.23
0.28
27.86
-0.22
41.85
0.43
11.10
0.28
31.80
-1.73
(% pred. precip.)
100.00
1.57
32.11
0.51
63.98
1.84
100.00
0.32
28.60
0.69
70.99
-0.59
100.00
1.02
26.53
0.64
75.96
-4.16
Difference
(cm)
-
0.30
1.75
-0.28
-9.77
8.00
.
-0.08
-4.53
-0.28
2.26
2.43
.
0.28
12.22
-0.28
-3.86
-11.91
(% meas. precip.)
-
0.56
3.26
-0.52
-18.19
14.89
.
-0.20
-11.59
-0.72
5.78
6.22
0.67
29.25
-0.67
-9.24
-28.51
Change in soil water storage.
DRAFT - DO NOT CITE OR QUOTE
4-26
-------
4.3.4 Test Plots in Live Oak, Georgia and Wenatchee, Washington
Of all the available studies, the one reported by Lane (1992), Khire (1995), and Khire et al.
(1997, 1999) is perhaps most interesting because of the scope and practical applicability of the
study to cover system analysis and design. The study involves field water balance evaluations
for three 30 m x 30 m cover system test plots at two landfills, one near Atlanta, Georgia ("Live
Oak") and the other near East Wenatchee, Washington ("Wenatchee"). The sites were selected
to represent humid and semi-arid climates, respectively. The Live Oak test plot has a cover
system with a 0.6-m thick CCL overlain by a 0.15-m thick vegetated silty topsoil layer. In
Wenatchee, one test plot has the same cover system as at the Live Oak site except that the CCL
is 0.6 m thick, and the other test plot models a capillary barrier consisting of a 0.75 m thick layer
of medium sand overlain by a 0.15-m thick silt topsoil layer. Climate, runoff, percolation, and
soil moisture data collected between 1992 and 1995 were reported by Khire (1995) and Khire et
al. (1997, 1999), and data collection is still ongoing as of 2002. Runoff and percolation is
collected in tanks and measured, while soil moisture content is measured by time domain
reflectrometry.
Khire (1995) and Khire et al. (1997) used their test plot data to assess the predictive capabilities
of the HELP and UNSAT-H models. The models were assessed by comparing model
predictions to measured hydrologic data for the three cover system configurations. The
predictions were performed using climatic data and laboratory-measured soil properties. Input
parameters that were not measured were estimated from published information. The input
parameters for this study were better defined than for most actual design projects. The UNSAT-
H predictions were conducted with a unit gradient lower boundary condition and a specified flux
upper boundary condition. Khire (1995) and Khire et al. (1997) drew the following conclusions
from their study:
• Properly simulating runoff is essential because the fraction of precipitation that is not
shed enters the cover system and may ultimately become percolation. Throughout most
of the monitoring period, HELP underpredicted runoff for the humid Live Oak site
(Figure 4-9) and overpredicted runoff for the semi-arid Wenatchee site with a CCL
(Figure 4-10). Overall, HELP underpredicted runoff by 740 mm (~ 90%) for the Live
Oak site and overpredicted it by 30 mm (~ 30%) for the Wenatchee site. Cumulative
runoff predictions made using UNSAT-H were reasonably accurate for the Live Oak site
(i.e., less than 3% error); however, season-to-season differences in runoff amounts were
significant. For the Wenatchee site, UNSAT-H underpredicted runoff by 50 mm (~
270%) for the plot with a CCL and predicted no runoff for the plot with a capillary
barrier. The underpredictions resulted in more water entering the soil in the simulations
than in the field. This resulted in higher soil water storage in the simulations than in the
field.
• Although HELP predicted ET fairly accurately for the Live Oak site, it was
underpredicted by only 70 mm (~ 4%), an accurate prediction of ET was not expected
given that more water entered soil due to the underprediction of runoff. Instead, an
overprediction of ET was expected unless the PET demand had already been met.
DRAFT - DO NOT CITE OR QUOTE 4-27
-------
UNSAT-H underpredicted ET for the Live Oak site by 300 mm (~ 15%). Examination of
the water-balance equation indicates that underpredicting runoff and fairly accurately
predicting ET, or vice versa, results in an overprediction of soil water storage and/or
percolation. Both HELP and UNSAT-H overestimated ET at the Wenatchee sites by
about 20 to 165 mm (~ 20 to 40%).
• HELP somewhat captured the trends in percolation at the Live Oak site, but
overpredicted total percolation by more than 700 mm (~ 300%) (Figure 4-11).
1992
1993
1994
1995
200
150 -
I
•£ 100 -
0>
o
0>
E
3
O
I I I I I I I I I I I
(a) Live Oak
I I I I I I I I I
Su Fa W Sp Su Fa W Sp Su Fa W Sp
92 92 93 93 93 93 94 94 94 94 95 95
I
T3
(D
O
"ro
o
CO
ro
o>
J5
30
25
20
15
10
5
-
-
,-|
|
r-
-,
•
D
PI
1
in
i
i
Measured
HELP
UNSAT-H
~L
ir
1
1
i
PI
1
I
i i
(b) Live Oak
-T
-i
|
-,
,-|
-
-
-
!• n
Su Fa W Sp Su Fa W Sp Su Fa W Sp
92 92 93 93 93 93 94 94 94 94 95 95
Figure 4-9. Measured and predicted cover system runoff at Live Oak site: (a) cumulative;
and (b) seasonal (from Khire, 1995).
DRAFT - DO NOT CITE OR QUOTE
4-28
-------
1992
1993
1994
1995
15
12
lz
O
>
is
^
E
3
O
i i i i i i i i r
(a) Wenatchee
HELP-
Measured-
UNSAT-H-
Fa W Sp Su Fa W Sp Su Fa W Sp
92 93 93 93 93 94 94 94 94 95 95
a
"H"
o
^ 6
T3
C
m
O
"ro
c
0
CO
ro 2
0
W
0
1 1
—
—
-
-
—
—
-
-
i
i i
1
• Measured
D HELP
D UNSAT-H
-i • •
1 1 1
i
(b) Wenatchee
—
—
JL ^
-
-
—
—
-
-
i i i i i i i i i r
Fa W Sp Su Fa W Sp Su Fa W Sp
92 93 93 93 93 94 94 94 94 95 95
Figure 4-10. Measured and predicted runoff for hydraulic barrier-type cover system at
Wenatchee site: (a) cumulative; and (b) seasonal (from Khire, 1995).
DRAFT - DO NOT CITE OR QUOTE
4-29
-------
1992
1993
1994
1995
120
_ Live Oak
/ UNSAT-H
I I I I I I I I I
Su Fa W Sp Su Fa W Sp Su Fa W Sp
92 92 93 93 93 93 94 94 94 94 95 95
Figure 4-11. Measured and predicted cover system percolation at Live Oak site (from
Khire, 1995).
1992
1993
1994
1995
41 I I I I I I I I I I
Fa
92
W Sp Su Fa W Sp Su Fa W Sp
93 93 93 93 94 94 94 94 95 95
Figure 4-12. Measured and predicted percolation for hydraulic barrier-type cover system
at Wenatchee site (from Khire, 1995).
DRAFT - DO NOT CITE OR QUOTE
4-30
-------
One reason why percolation was overpredicted is that there was additional water in the soil
caused by the underprediction of runoff. Another factor that contributed to the overprediction of
percolation is the unit hydraulic gradient used by HELP to model unsaturated vertical flow.
HELP assumes that water in the soil flows vertically downward under a unit hydraulic gradient
(i.e., hydraulic gradient = 1). Khire (1995) and Khire et al. (1997) indicate that the hydraulic
gradient in the field rarely equaled "1" and, for most of the time, was oriented vertically upward.
UNSAT-H underpredicted percolation for the Live Oak site only slightly, by about 60 mm. Both
HELP and IMSAT-H underpredicted percolation for the Wenatchee site with a CCL barrier
(Figure 4-12). However, at least part of this difference is believed to have been caused by the
preferential flow of water and snow melt through cracks and animal burrows in the winter of
1995. Prior to that time, both models had overpredicted percolation. UNSAT-H significantly
overpredicted percolation for the Wenatchee site with a capillary barrier (Figure 4-13). One
reason why percolation was overpredicted by over 90 mm (~ 2,000%) is that there was
additional water in the soil caused by the underprediction of runoff.
1992
1993
1994
1995
15
E 12
o
o
0>
a.
0>
m
3
E
3
O
T
T
T
Wenatchee:
Capillary Barrier
Fa W Sp
92 93 93
Figure 4-13. Measured and predicted percolation for capillary barrier-type cover system
at Wenatchee site (from Khire, 1995).
DRAFT - DO NOT CITE OR QUOTE
4-31
-------
4.4 Recommendations for Application of Water Balance Models
The specific water balance analysis method and input parameters to use for analysis and design
of a cover system should be selected based on the purpose of the analysis and project-specific
factors such as climate, type of cover (i.e., hydraulic barrier, ET barrier, or capillary barrier), and
cover system components. Given the inconsistencies in water balance analysis results (e.g., the
models sometimes overpredict and sometimes underpredict the various components of the water
balance), uncertainties in soil properties and long-term barrier integrity (e.g., CCL hydraulic
conductivity may increase over time if the CCL is not adequately protected), and other factors,
significant engineering judgment must be applied when performing a water balance analysis for
a specific site. The following general recommendations are made regarding the use of water
balance methods for the design of cover systems:
• Percolation rates through cover systems with GM, GM/CCL, or GM/GCL hydraulic
barriers should be very low when these barriers are properly constructed due to the
effectiveness of these barrier types in preventing water migration through the barrier.
Both the simplified manual method and the HELP model are well suited to performing
analyses to demonstrate the effectiveness of these type of barriers in minimizing
percolation.
• Estimated percolation rates through hydraulic barriers layers containing GMs for various
categories of annual rainfall were provided by Gross et al. (1997) (Table 4-6). These
estimates can be used by design engineers as a check of percolation rates calculated on a
project-specific basis using either the simplified manual method or the HELP model.
Percolation rates were calculated by Gross et al. (1997) using the HELP model with
synthetic rainfall data generated by the model for several different cities in each rainfall
category and the following ranges of input parameters: (i) fair grass vegetation; (ii) sandy
loam and silty clay loam topsoil; (iii) 5 and 20% cover system slopes; (iv) coarse sand
and GN internal drainage layers; and (v) 10-year synthetic weather records.
Table 4-6. Percolation Rates through Cover Systems with Barriers Incorporating GMs
Estimated Using the HELP Model (from Gross et al., 1997).
Average Annual
Rainfall (mm)
100-300
300-600
600-800
800-1,000
1,000-1,600
Average Percolation Rates
(mm/yr)
GM Barrier
0-0.05
0.002-0.3
0.1-1
0.3-2
1-5
GM/CCL or GM/GCL Barrier
0-0.005
0.0002-0.03
0.01-0.1
0.03-0.2
0.1-0.5
DRAFT - DO NOT CITE OR QUOTE
4-32
-------
• Either the simplified manual method or the HELP model can be used for the design of
internal drainage layers underlain by hydraulic barriers containing a GM. A discussion
of the design storm to use with each method is given below.
• Neither the simplified manual method nor HELP are capable of serving as a water
balance predictive tool using estimated or default input data. The HELP model has
limited capability as a predictive tool when calibrated using site-specific data.
• Any of the water balance analysis methods may be used for evaluating percolation
through cover systems with CCL or GCL hydraulic barriers. While methods that
incorporate unsaturated flow models are potentially more accurate than methods where
saturated conditions are assumed for flow through the hydraulic barrier, the latter
methods (i.e., simplified manual method and HELP model) are easier to use. These latter
methods are likely to overpredict actual percolation rates for humid sites.
• For capillary-barrier and ET-barrier cover systems, a water balance analysis method that
can correctly model unsaturated flow is preferred. Thus, LEACHM, UNSAT-H,
SoilCover, or HYDRUS-2D is preferable to the HELP model for evaluation of these
types of systems.
• For cover systems in any climate that rely on enhanced ET to minimize percolation,
methods that correctly model unsaturated flow and that allow different vegetation
scenarios to be input, such as LEACHM, UNSAT-H, SoilCover, or HYDRUS-2D, are
preferred.
• Reference should be made to the available technical literature for the best available
information on the tendencies of the various water balance models to either underpredict
or overpredict the various components of the water balance for both wet and arid climatic
conditions. This information should be considered in interpreting the results of project-
specific water balance analyses.
• Reference should be made to the technical literature for new models that may be
developed in the future with enhanced capabilities for the performance of cover system
water balance analyses. All of the available models have their strengths and weaknesses.
There remains room for improvement of the models and their specific applications
• Due to the difficulty in performing accurate analytical water balances, field water
balances should be performed, whenever possible, to verify the analytical results. This is
especially the case for alternative cover systems.
• An important input parameter in the design of cover system internal drainage layers for
hydraulic barrier cover systems is rainfall intensity and duration. As previously
discussed, the HELP model is limited to using daily rainfall data, and this does not
capture short-term intense peaks in storm events. Koerner and Daniel (1997) have
suggested that hourly rainfall data be considered along with the simplified manual
method to calculate percolation through the cover soil into the internal drainage layer
(PERC*). They presented an example calculation of the sensitivity of PERC* to the use
of monthly, daily, or hourly precipitation data. The example assumes a site near Austin,
DRAFT - DO NOT CITE OR QUOTE 4-33
-------
Texas, with a 200-m long 3H: IV slope and a surface runoff coefficient of 0.4. The
results of their analysis were as follows:
o PERC* = 0.011 mm/hr, using the simplified manual method with the average
monthly temperature, duration of sunlight, and precipitation data from Austin;
o PERC* = 1.3 mm/hr using the HELP model with historical daily precipitation
data from 1974-1977 for San Antonio and all other climatic data generated for
Austin; and
o PERC* = 50 mm/hr using Eq. 4.7 with the probable maximum 6-hr precipitation
event for the project vicinity (i.e., 500 mm).
• Koerner and Daniel (1997) noted that the calculated peak flow rate based on hourly storm
data is more than one order of magnitude larger than the calculated peak flow based on
daily precipitation values. Because of this, they recommended that hourly precipitation
data be considered to conservatively calculate peak flow rates into the drainage layer and
to determine if the drainage layer has adequate capacity to transmit the peak flow during
extreme storm events.
For this guidance document, PERC* was calculated for the same example as above using the
HELP model with climatic data generated synthetically for Austin for a 20-year simulation
period. For the authors' simulation, PERC* = 3.1 mm/hr. This calculated PERC* is about
2.5 times larger than the value obtained by Koerner and Daniel (1997) of 1.3 mm/hr using
the historical weather data for 1974-1977 for San Antonio. This result reinforces the
comment made previously in this chapter that the HELP precipitation database for the period
1974-1977 reflects unusually dry weather for certain parts of the U.S. More generally, short-
duration rainfall records may not contain wet weather cycles or intense storm events that
control design. Also, as Koerner and Daniel (1997) noted, the rate of infiltration into a cover
system soil will be limited by the hydraulic conductivity of the cover soil materials. If it is
assumed in the above example that the cover soil has a saturated hydraulic conductivity of
1 x 10"6 m/s, then from Eq. 4.8, the maximum possible rate of infiltration into the cover for a
non-ponded surface condition is 3.6 mm/hr , approximately the rate of percolation calculated
with the HELP model and daily rainfall data generated synthetically for Austin, Texas, (i.e.,
3.1 mm/hr). Thus, for typical cover systems with low to moderately permeable surface and
protection layers, it will often be adequate to use the HELP model and a synthetic rainfall
record with a sufficiently long simulation period (e.g., 20 years) to calculate lateral drainage
and hydraulic head. Alternatively, Eq. 4.8b can be used directly to obtain a conservative
value of PERC* for design.
4.5 Design of Drainage Layers
4.5.1 Simplified Manual Method
The required hydraulic properties of the cover system drainage layer are a function of the
expected peak rate of percolation into the drainage layer (PERC* in Sections 4.2 and 4.3), the
length of the cover system slope, the inclination of the cover system slope, and other factors.
DRAFT - DO NOT CITE OR QUOTE 4-34
-------
Assuming no change in water storage in the drainage layer material, lateral flow in that layer is
equal to percolation through the cover soil into the layer (PERC*) minus percolation through the
hydraulic barrier (PERC). From Eq. 4.4:
L = PERC* - PERC (Eq. 4. 14)
where all terms are as defined previously. Assuming steady-state conditions, the maximum flow
in the drainage layer is given by Eq. 4.5, repeated here:
£L i (PERC *- PERC)
8.64x10'
where: qm = maximum flow rate in drainage layer per unit width perpendicular to the direction of
flow (mVs/m); £ = slope length (m); and other terms are as defined previously. For design of
drainage layers, PERC can be conservatively assumed to be zero: that is, all percolation through
the cover soil (PERC*) is assumed to become lateral flow in the drainage layer. For this case:
(Eq.4.!5)
»
8.64 xlO7
The hydraulic transmissivity of the drainage layer must be adequate to accommodate this flow.
In the simplified manual method, the DuPuit-Forcheimer assumptions are used along with the
further assumption that the line of seepage is parallel to the cover system slope to calculate the
required drainage layer hydraulic transmissivity. For these assumptions, the hydraulic gradient
is constant and equal to the sine of the slope angle:
i = (sinp) (Eq. 4.16)
where |3 = slope angle (degrees). The required hydraulic transmissivity of the drainage layer is
then obtained using Darcy's equation and the known values of qm and i:
0h = (qm/i)FS (Eq.4.17)
Substituting Eqs. 4.15 and 4.16 into Eq. 4.17 results in:
/ PERC *
IJ^KI - F§ (Eq.4.18)
8.64 xlO7 ship
where: 9h = required hydraulic transmissivity of drainage layer (m2/s); FS = factor of safely
(dimensionless); and other terms are as defined previously. As previously discussed in Section
2.4.2.3, a minimum FS value of 2 is recommended for cases where the uncertainty in input
parameters is low and the consequences of failure are small. For many situations, a larger FS
DRAFT - DO NOT CITE OR QUOTE 4-35
-------
may be appropriate. Koerner and Daniel (1997) have recommended using a FS value of at least
5 to 10 to account for uncertainities in the hydraulic conditions.
The maximum hydraulic head in the drainage layer for the assumptions given previously is:
(Eq. 4.19)
where: hm = maximum hydraulic head (m); kd = hydraulic conductivity of drainage layer (m/s);
and qm is as defined previously. The maximum hydraulic head for this set of assumptions occurs
at the base of the slope. The required thickness (measured perpendicular to the slope) of the
internal drainage layer is obtained from the equation:
= (hm/cosp)FS =
(Eq. 4.20)
where: tm = the required thickness of the internal drainage layer (m); and other terms are as
defined previously. The actual thickness of the internal drainage layer must be larger than tm in
order for pressure head not to build up in the layer. The definition of the thickness, head, and
depth of flow on a slope is shown in Figure 4-14.
h = t cos P
= dcos|3
Figure 4-14. Definition of liquid depth (d), thickness (t), and hydraulic head (h), above a
hydraulic barrier.
DRAFT - DO NOT CITE OR QUOTE
4-36
-------
4.5.2 Refinement to Simplified Manual Method
For a sloping drainage layer receiving a constant rate of percolation (PERC*), flow in the layer
is not actually parallel to the slope as assumed in the previous subsection. Rather, as the
hydraulic head builds up on the slope, the phreatic surface takes on a curved shape. Figure 4-15
illustrates this condition for a cover system slope with a toe drain. For this condition, the
hydraulic gradient is not constant but varies along the slope length.
Internal Drainage Layer
Hydraulic Barrier
Drain
Figure 4-15. Hydraulic head distribution on a cover system slope with a toe drain.
An improved estimate of maximum hydraulic head in the internal drainage layer that takes
account of the varying hydraulic gradient (while maintaining use of the DuPuit-Forcheimer
assumptions) can be obtained using the equations from Giroud et al. (1992b) and Giroud and
Houlihan (1995):
hm=(j^cosp/2)
tan2p
4 PERC *
kcosp
.1/2
-tanp
(Eq. 4.21)
where all terms are as defined previously, and j is given by Eq. 4.21:
j = l-0.12exp[-(log(8X/5)5/8)2]
(Eq. 4.22)
DRAFT - DO NOT CITE OR QUOTE
4-37
-------
where:
PERC*
k tan P sin P
(Eq. 4.23)
It is noted that Eq. 4.20 tends to the simplified solution of Eq. 4.18 when PERC*/k tends towards
zero and/or |3 is very large. Values of average hydraulic head, havg (m), for a given value of hm
can be obtained from Figure 4-16. For the case of (PERC*/k cos|3) < 0.25 tan2|3:
PERC * £
avs~2ksinpcosp
(Eq. 4.24)
It is suggested that for design of internal drainage layer, hm be used from single storm event
analyses to size the drainage layer. In contrast, it is suggested that havg be used to calculate long-
term PERC values. For the simplified manual method, PERC* to calculate hm should be derived
using hourly water balance calculations for the design storm (limited by Eq. 4.8 as previously
discussed) and PERC* to calculate havg should be derived using monthly water balance
calculations.
1.0
0.8
0.6
0.4
0.2
0.0
io~4 10'3 io~2 10 ~1 10° icr1
Dimensionless Factor
10'2 10'
,-3
10"
Figure 4-16. Dimensionless factor for calculating (have/hm) for internal drainage layers.
(from Giroud and Houlihan, 1995).
4.5.3 HELP Model
In the HELP model, lateral drainage in internal drainage layers is modeled by an analytical
approximation to the steady-state solution of the Boussinesq equation (Darcy's equation coupled
with the continuity equation), employing the Dupuit-Forcheimer assumptions. Hydraulic heads
calculated for internal drainage layers in the HELP model are similar to those that would be
calculated using the equations presented by Giroud and Houlihan (1995) for equal values of
PERC*. Based on the example calculation in Section 4.4 of this document, the HELP model can
be used directly for calculating lateral flow and hydraulic heads in cover system internal
DRAFT - DO NOT CITE OR QUOTE
4-38
-------
drainage layers. However, in using the model, the user should select a weather data generating
option that produces extreme wet weather periods for the project site. Use of the 1974-1977
HELP model internal weather database will not typically be adequate.
4.6 Design of Slope Transitions
Design of internal drainage layers at benches and other slope transitions is critical to the
effective functioning of the drainage layer. If not properly designed, flow will back up and
generate hydraulic pressure at the slope transition. For flow not to back up in a drainage layer
flowing full, flow capacity (q) across the slope transition must not decrease. Flow capacity for
laminar flow parallel to a slope is equal to the hydraulic gradient multiplied by the hydraulic
transmissivity of the drainage layer material. This design requirement is illustrated in Figure 4-
17.
Figure 4-17. Continuity of flow across a slope transition for laminar porous media
condition.
For many conventional cover system designs, the hydraulic gradient on the flatter part of the
slope transition will be about one order of magnitude lower than the hydraulic gradient on the
steeper part. For example, the gradient of a 3H: IV slope is 0.32, whereas the gradient reduces to
0.03 for a 3% slope inclination typical of a cover system bench. For this condition, to prevent
backup of flow and build-up of hydraulic head for drainage layer flowing full, the hydraulic
transmissivity of the drainage layer on a cover system bench or slope transition will need to be
about one order of magnitude larger than that of the drainage layer on the sideslope.
More generally, based on Figure 4-17, the slope transition should be designed such that:
0h2 > 0hi (sinpi/sinp2) (Eq. 4.25)
where all terms are as defined previously. The subscript 1 refers to the portion of the drainage
layer on the steeper upslope side of the transition, and the subscript 2 refers to the drainage layer
DRAFT - DO NOT CITE OR QUOTE
4-39
-------
on the flatter downslope side of the transition (Figure 4-17). Eq. 4.25 can be used directly to
analyze and design geosynthetic drainage layers for which hydraulic transmissivities are either
known or measured in the laboratory. For granular drainage materials where materials are
typically specified in terms of a required hydraulic conductivity and thickness, Eq. 4.25 is recast
as:
k2 > ki (tmi/tm2) (sinpi/sinp2) (Eq. 4.26)
where all terms are as defined previously. For Eq. 4.25 to be valid, tmi and tm2 must be less than
the total thickness of the drainage layer.
The concept of having a larger internal drainage layer hydraulic transmissivity (or hydraulic
conductivity) on a slope bench compared to the adjacent upslope portion of the cover is
illustrated in Figure 4-18(a). This approach is conveniently achieved with geosynthetic drainage
layers; it is more difficult to implement with granular drainage materials because it requires very
coarse-grained materials on the benches or slope transitions while meeting filter criteria at the
interface between drainage materials. Other options for designing benches and slope transitions
are shown in Figures 4-18(b), (c), and (d). These include:
• installing a perforated pipe within the slope transition to convey water to outlet pipes
(Figure 4-18(b)); this approach is technically acceptable, but there can be a problem with
the pipes freezing and plugging; also, it is essential that the pipes remain open and not be
plugged or damaged by maintenance personnel; in addition, the discharge from the pipes
may tend to erode soil beneath the pipes, and the surface should be adequately protected
to prevent excessive erosion;
• installing a perforated pipe within the slope transition to convey water to a downdrain or
downchute; this has the advantage of keeping the piping below the surface, where it can
be protected from freezing; because the surface of the bench is normally sloped to
provide surface drainage, the perforated pipe can follow the slope of the bench and
provide gravity drainage to the outlet point; the outlet must still be protected and cannot
be obstructed or clogged; and
• allowing the drainage layer to daylight on the bench. The bench must be suitably
protected to prevent erosion; also, the outlet cannot freeze, which makes this approach
questionable in northern climates.
DRAFT - DO NOT CITE OR QUOTE 4-40
-------
(a)
Dover Soil
Sand
Drainage Layer
Grave
Increase Hydraulic
Transmissivity
(b)
Cover Soil
Perforated Pic
Drainage Pipe
Gravel for
Erosion Protection
(C)
- Cover Soil
Sand
Drainage Layer
Perforated Pipe-
(Convey water to down drain or downchute)
(d)
Dover Soil
Surface Collection System
(Convey water to down drain or downchute)
Sand
Drainage Layer
Figure 4-18. Design options for cover system slope transitions.
DRAFT - DO NOT CITE OR QUOTE
4-41
-------
4.7. Design of Filter Layers
4.7.1 Overview
To prevent clogging of internal drainage layers, it is often necessary to install a granular or GT
filter layer directly over the drainage layer material. Several of the cover system slope stability
problems described in Chapter 7.4 of this document were due, at least in part, to inadequate filter
layer design. The function of the filter in cover system applications is to limit the migration of
fines from the overlying cover soil into the internal drainage layer, while allowing unimpeded
percolation from the cover soil into the drainage layer. If the drainage material is a granular soil,
the filter material may be either soil or GT. If the drainage material is itself a geosynthetic, the
filter layer will also need to be a GT.
Filter criteria establish the relationship of grain sizes necessary to retain adjacent materials and
prevent clogging of a drainage layer, while allowing unimpeded percolation. Criteria for the
design soil and GT filter layers are discussed below.
4.7.2 Soil Filters
Soil filters usually consist of fine to medium sands when placed over coarse sand or fine gravel
drainage layers. The filter particle size distribution must be carefully selected. Fortunately,
there is a considerable body of information available to use in selecting a filter particle size
distribution (Koerner and Daniel, 1997). Typically, the criteria described in Cedergren (1989)
are used. To prevent piping from the overlying cover soil into the filter layer, and from the filter
into the drainage layer, these criteria require, respectively:
Dis (filter)/D85 (cover soil) < 4 to 5 (Eq. 4.27)
and:
DIS (drainage layer)/D85 (filter) < 4 to 5 (Eq. 4.28)
To maintain adequate permeability of the filter layer and drainage layer, these criteria require,
respectively:
Dis (filter)/Di5 (cover soil) > 4 to 5 (Eq. 4.29)
and:
Dis (drainage Iayer)/Di5 (filter) > 4 to 5 (Eq. 4.30)
where: Dgs = particle size at which 85% by dry weight of the soil particles are smaller (mm); and
DIS = particle size at which 15% by dry weight of the soil particles are smaller (mm). The
criteria should be satisfied for all layers or media in the drainage system, including cover soil,
filter material, and drainage material.
4.7.3 GT Filters
A GT must be installed over a GN or drainage core when the overlying material is to be a cover
soil. The primary function of the GT in this application is as a filter layer. As with soil filter
layers, GT filters must allow percolation from the cover soil to pass unimpeded into the drainage
layer while retaining the cover soil and limiting the migration of particles from the cover soil.
DRAFT - DO NOT CITE OR QUOTE 4-42
-------
As with soil filters, the design of GT filters involves a two-step process: first to assess
permeability (or permittivity) and second to evaluate soil retention (or apparent opening size).
The first step in design of a GT filter is to establish the GT permittivity (*F) requirements. The
usual formulation involves expressing the minimum allowable GT permittivity (^min) as a
multiple of the required permittivity (*Freq) to maintain flow continuity from the cover soil, as
follows:
ymm=FSyreq (Eq.4.31)
and:
¥ = -^ (Eq. 4.32)
where: Y = GT permittivity (s"1); kn = GT cross-plane hydraulic conductivity (m/s); and t =
thickness of GT at a specified normal pressure (m). A minimum FS of 5 to 10 is recommended.
The testing of a GT for permittivity is conceptually similar to the testing of granular soils for
permeability. In the U.S., the testing is usually performed using the permittivity test, ASTM D
4491. Alternatively, some design engineers prefer to work directly with permeability and
require the GT's hydraulic conductivity to be some multiple of the adjacent soil's hydraulic
conductivity (e.g., 5 to 10, or higher).
The second step of the design of a GT filter is intended to assure adequate retention of the cover
soil. There are several methods available for establishing the soil retention requirements of GT
filters. Most of the available approaches, as applied to a cover system, involve a comparison of
the cover soil particle size characteristics to the 95% opening size of the GT (i.e., defined as 095
of the GT). The 095 is the approximate largest soil particle size that can pass through the GT.
Various test methods are used to estimate 095: (i) in the U.S., wet sieving is used and the value
thus obtained is called the apparent opening size (AOS), ASTM D 4751; (ii) in Canada and some
European countries, hydrodynamic sieving is used and the value thus obtained is called the
filtration opening size (FOS); and (iii) in other European countries, wet sieving is used.
The simplest of the available design methods involves a comparison of the GT AOS to standard
soil particle sizes as follows (Koerner, 1998):
• for soil with < 50% passing the No. 200 sieve (0.074 mm): 095 < 0.59 mm (i.e., AOS of
the GT > No. 30 sieve); and
• for soil with > 50% passing the No. 200 sieve: 095 < 0.33 mm (i.e., AOS of the GT > No.
50 sieve).
Alternatively, a series of direct comparisons of GT opening size (095, 05o, or Oi5) can be made to
some soil particle size to be retained (D90, D8s or Di5). The numeric value depends on the GT
type, soil type, flow regime, etc. For example, Carroll (1983) recommends the following
relationship:
095 < (2 or 3)D85 (Eq. 4.33)
DRAFT - DO NOT CITE OR QUOTE 4-43
-------
where: Dgs = particle size at which 85% by dry weight of the soil particles are smaller (mm); and
095 = the 95% opening size of the GT (mm). As shown by Giroud (1992, 1996), Eq. 4.33 should
only be used if the coefficient of uniformity of the soil to be protected is less than four. General
procedures, applicable for all values of the coefficient of uniformity of the soil to be protected,
are available: see Giroud (1982), Lafleur et al. (1989), and Luettich et al. (1992).
Occasionally, a drainage layer is placed directly against a GCL. For GT-encased GCLs, the GT
components may not be adequate to prevent migration of bentonite into the drainage layer. The
required filter criteria for this condition are under study, and the manufacturer's and technical
literature should be consulted. One study indicated that a 350 g/m2 nonwoven, needlepunched
GT provided adequate protection from bentonite migration for all GCLs investigated (Estornell
and Daniel, 1992).
DRAFT - DO NOT CITE OR QUOTE 4-44
-------
Chapter 5
Gas Emission Analysis and Collection System Design
5.1 Introduction
This chapter provides information on select topics related to cover system gas emission analysis
and collection system design. The specific topics discussed in this chapter are:
• mechanisms of gas generation and emission (Section 5.2);
• characteristics of selected gas emissions models (Section 5.3); and
• design of gas collection systems (Section 5.4).
5.2 Mechanisms of Gas Generation and Emissions
5.2.1 Overview
Landfill gas (LFG) is the byproduct of anaerobic decomposition of the organic material
placed in a landfill during its active life. Landfill gas emissions may create hazardous
situations. The nature and extent of the hazard depends on the emission rate, toxic constituent
concentration, and the relative concentration of the flammable components. Human health and
the environment may be adversely affected because of its potential to: A) create
flammable/explosive conditions within enclosed spaces; B) contain a mixture of toxic and
hazardous air pollutants (HAPs); C) contain large quantities of greenhouse gases (carbon
dioxide, methane, nitrous oxide); D) contain volatile organic compounds (VOC) that are
precursors to the formation of ozone and smog; and E) have an odiferous smell that is
objectionable to many people. The literature indicates that LFG may contain more than 100
non-methane organic compounds (NMOC's) (EPA 1997 a and b). Over thirty of the NMOC's
are classified as HAPs (EPA 1997 a and b). Landfills are listed as a source in EPA's Urban Air
Toxic Strategy and have been identified for residual risk evaluation.
The LFG emission rate through the cover system of a landfill is dependent on the gas
generation rate, whether the facility has a liner system, site hydrogeology, characteristics of the
cover system, and characteristics of any gas control system. The Landfill Gas Emissions Model
(LandGEM) estimates landfill gas emissions based on the age of the landfill, the quantity of
waste placed within it, waste acceptance rate, and other site specific information. LandGEM
uses a first-order decomposition rate equation to make the estimates (Thorneloe, 1999). A
personal computer-based version of LandGEM can be downloaded from EPA's website at
http://www.epa.gov/ttn/catc/products.htmltfsoftware. A user's manual is also available on this
website. The software has various sets of defaults values that can be adjusted by the user. One
set is for those sites where the CAA requirements are determined to be applicable and
appropriate. The other set is typically used for emission inventories and is less conservative
than the CAA defaults. Site-specific data can also be used if available. EPA is also developing
GUIDANCE FOR EVALUATING LANDFILL GAS EMISSIONS FROM CLOSED OR
ABANDONED FACILITIES under EPA Contract number 68-C-00-186 Task Order Number 3.
5-1
DRAFT - DO NOT CITE OR QUOTE
-------
This guidance when published will provide procedures and a set of tools for evaluating the
nature, extent, risks and hazards associated with LFG emissions to ambient air, LFG subsurface
vapor migration due to landfill gas pressure gradients, and subsurface vapor intrusion into
buildings. Figure 5.1 provides a flow diagram of the guidance.
Conduct ' "Hot Spot"
Ho
Jrooram
Figure 5-1. Flow Diagram of Guidance for Evaluating Air Pathway at Older, Closed Landfills
MSW landfills constructed or operated after October 9, 10993 are governed by the RCRA
Subtitle D regulations. These regulations establish siting restrictions, and design, operating,
and monitoring standards that are designed to minimize the potential for environmental damage.
Additionally, rules and regulations implementing the Clean Air Act (CAA) establish Maximum
Achievable Control Technology (MACT) standards that are applicable to landfills that exceed
size and age threshold. The CAA regulations require landfill gas collection and control systems
to be installed at landfills that (1) contain at least 2.5 million megagrams (Mg) or 2.5 million
cubic meters of waste and (2) emit 50 Mg per year or more of NMOCs (EPA, 1998). EPA's
emission guidelines (EGs) apply to existing landfills that were in operation from November 8,
1987 to May 30, 1991. EPA's new source and performance standards (NSPS) apply to any
existing landfill constructed on or after May 30, 1991 or which undergo changes in design
capacities on or after May 30, 1991. To help evaluate performance of the gas collection and
control system, the NSPS/EGs (60 CFR §753) require that:
"Each owner or operator of a MSW landfill gas collection and control system used to comply
with the provisions of Sec. 60.752(b)(2)(ii) of this subpart shall... (d) Operate the collection
system so that the methane concentration is less than 500 parts per million above background
at the surface of the landfill. To determine if this level is exceeded, the owner or operator shall
conduct surface testing around the perimeter of the collection area along a pattern that
traverses the landfill at 30 meter intervals and where visual observations indicate elevated
5-2
DRAFT - DO NOT CITE OR QUOTE
-------
concentrations of landfill gas, such as distressed vegetation and cracks or seeps in the
cover...."
For further information on the requirements of these regulations and available technical
documents and fact sheets, refer to http://www.epa.gov/ttnatw01/landfill/landflpg.html.
Most hazardous waste landfills do not include significant quantities of garbage and other
biodegradable materials. Hence the ability of a hazardous waste landfill to generate LFG is
severely limited by the lack of carbon, nutrients, and moisture. Additionally, the Land Disposal
Restriction (LDR) rules (see 40 CFR 268.7) have required generators of hazardous waste to
meet the best demonstrated available technology (BDAT) standards since 1986. The LDR rules
require generators to meet BDAT standards before any hazardous waste is placed in a landfill.
The BDAT standards are designed to substantially diminish the toxicity of the waste or
substantially reduce the likelihood of migration of hazardous constituents from the waste. The
BDAT standards are waste code and media specific but in all cases the maximum allowable
constituent concentration prior to disposal is measured in the parts per million or less range.
Hazardous waste landfills are also required to install covers that are hydraulically impermeable.
The existence of a hydraulically impermeable geomembrane (see Section 1.2.2) also limits
uncontrolled LFG emissions from these types of facilities. There are, however, no regulatory
requirements regarding LFG emissions from hazardous waste landfills. In contrast, hazardous
waste disposed prior to 1986 may have included highly concentrated materials. Additionally,
prior to 1980 most landfills were closed with a cover system consisting primarily of earthen
materials that were permeable by design. . Emission of particulates from waste containment
facilities and remediation sites is also a concern for the Agency. However, this type of emission
is not addressed in this guidance document.
5.2.2 MSW Landfill Gas Generation
The anaerobic decomposition of MSW produces two principal gases, methane (CH4) and carbon
dioxide (€62), and much smaller quantities of other gases, including nitrogen, oxygen, sulfides,
ammonia, and other constituents, and trace amounts of a variety of NMOCs, typically including
vinyl chloride, ethylbenzene, toluene, and benzene (Tchobanoglous, 1993; EPA, 1997a, 1997b).
The typical constituents found in MSW landfill gas and their concentrations are listed in Table
5-1. Though it is not included in Table 5-1, landfill gas is also typically saturated with water
vapor at levels of 1 to 5% by volume. Typical concentrations of NMOCs in landfill gas are
presented in Table 5-2.
The methane in MSW landfill gas can accumulate in enclosed or confined spaces (typically near
the perimeter of the landfill, but in some cases at considerable distances from the perimeter) in
concentrations that are odorous, asphyxiating, toxic, corrosive, flammable, or even explosive.
Besides methane, other gas constituents, such as hydrogen gas, are also explosive, and certain
gas constituents, such as hydrogen sulfide gas, are also toxic at a certain concentrations.
Landfills have been identified as the source of nearly 30 HAPs, including the constituents listed
in Table 5-2. A more comprehensive list of the different NMOCs and HAPs in landfill gas is
contained in EPA's AP-42 which provides guidance for estimating landfill gas emissions (EPA
5-3
DRAFT - DO NOT CITE OR QUOTE
-------
1997a and b). The source of these HAPs is primarily household and small quantity generator
hazardous wastes, including paints, solvents, pesticides, and adhesives.
Because of the hazards posed by MSW landfill gas, care must be taken in handling the gas.
CAA regulations establish requirements for MSW landfill gas collection and control at certain
facilities, as described in Section 1.4.
Table 5-1. Typical landfill gas constituents (from Tchobanoglous et al., 1993).
Constituent Percent by Volume
Methane 40-60
Carbon dioxide 40-60
Nitrogen 2-5
Oxygen 0.1-1
Ammonia 0.1-1
Sulfides, disulfides, mercaptans, etc. 0-0.2
Hydrogen 0-0.2
Carbon monoxide 0-0.2
Trace constituents 0.01-0.6
Table 5-2. Typical concentrations of NMOCs in gas from 25 landfills in southern
California (from Pierce et al., 1998).
Trace Gas Constituent Average Constituent Concentrations
(ppm by volume)
Benzene
Chlorobenzene
1,1-Dichloroethane
1,2-Dichloroethane
Methylene Chloride
Tetrachloroethene
Tetrachloromethane
Toluene
1,1,1-Trichloroethane
Trichloroethene
Vinyl Chloride
Range
0.432-21.8
0.054-5.24
0.10-15.9
0.01-3.74
0.10-56.3
0.30-28.2
0.001-0.413
8.37-67.7
0.012-8.28
0.293-13.6
0.277-16.8
Mean
2.76
0.606
2.39
0.29
10.5
3.24
0.046
28.3
0.715
1.60
1.99
Landfill gas generation is often considered to occur in five sequential phases, as shown in
5-4
DRAFT - DO NOT CITE OR QUOTE
-------
Figure 5-1. During Phase I, the initial adjustment phase, waste placement starts, and the waste
begins to accumulate moisture. Microbes in the waste begin to acclimatize to the landfill
environment. With plenty of substrate and nutrients available, aerobic microbes start to degrade
the waste, producing water, carbon dioxide, organic acids, and inorganic minerals. The aerobic
decomposition is sustained by the oxygen trapped within the waste mass. Because Phase I is
relatively short lived and involves aerobic decomposition, it is sometimes combined with Phase
II and referred to as the "aerobic phase".
During Phase II, the transition phase, oxygen trapped within the landfill is depleted and the
landfill transitions from an aerobic to anaerobic environment. Since the amount of trapped
oxygen is limited, this stage is also relatively short lived (i.e., a few days to a few months). As
oxygen is depleted, a trend for reducing conditions is established, with a shifting of electron
acceptors from oxygen to nitrates and sulfates. Reduction of these latter molecules, often
produces nitrogen gas and hydrogen sulfide gas. In addition, the carbon dioxide level begins to
increase, causing the formation of carbonic acid and a decrease in the leachate pH to the acidic
range. Waste temperatures are hottest during this phase, reaching 54 to 71°C.
In Phase III, the acid phase, waste is degraded anaerobically. The waste first undergoes
hydrolysis, where larger organic molecules are converted into shorter, soluble molecules and
hydrogen gas is produced. Acidogenic bacteria then convert the hydrolyzed compounds into
volatile organic acids (VOAs). The acids, in turn, cause the pH to drop (e.g., to 5.5 to 6.5) in
1.0
o
3
£
0.
£ 0.6
re
O
re °-4
-------
turn causing heavy metals concentrations to rise in the leachate. Viable biomass growth
associated with the acidogenic bacteria and the rapid consumption of substrate and nutrients are
the predominant features of this phase. The primary gas formed during this stage is carbon
dioxide.
In Phase IV, the methane fermentation phase, the VOAs and hydrogen gas produced by the
acidogens are converted into methane by methanogenic bacteria. Both acid production and
methane fermentation occur during this phase; however, methane fermentation predominates.
The highest landfill gas generation rates occur during this phase. As the VOAs are utilized, the
pH of the leachate increases to more neutral values (e.g., 6.8 to 8) and heavy metals
concentrations decrease. Sulfates and nitrates are reduced to sulfides and ammonia. Gas
temperatures have dropped by this phase to about 38 to 54°C. Gas production probably begins
to drop off at the lower end of this temperature range. As described by Hutric and Soni (1997),
a study of an experimental MSW digestor showed that gas generation rates peaked at two
temperatures: about 40 °C, when mesophilic bacteria are present, and between 55 and 60°C,
when thermophilic bacteria are present. At temperatures below 40 °C, gas generation rates
decrease rapidly with decreasing temperature.
By Phase V, the maturation phase, the landfill has matured and the readily biodegradable
material has been stabilized (i.e., converted to methane or carbon dioxide). Biodegradation is
limited by lack of readily degradable substrate and nutrients, so biological activity slows. The
landfill gas production rate, consequently, also decreases. Both carbon dioxide and methane
gases are produced, but at much lower rates than in Phase IV. Towards the latter part of this
phase, the landfill may become aerobic, with oxidizing conditions, and small amounts of
oxygen and nitrogen gases may be present.
Since landfills are heterogeneous and all waste is not placed at the same time, the stages described
above typically occur concurrently in different areas and depths of an active or recently closed
landfill. The dichotomy between stages is often masked when a landfill is active and new waste is
being added to old. After a landfill closes, the landfill tends to move into Phase IV, with the newer
waste just keeping the landfill at this phase for a longer time period.
The rate of waste degradation is controlled by the amount and type of degradable materials in
the waste, waste temperature, waste moisture content, and other factors. Food waste may
degrade about five times faster than yard waste, fifteen times faster than paper, and fifty times
faster than wood or leather. Degradation is enhanced (the reaction rates increase) by the initial
temperature increase caused by the heat released from aerobic degradation. The temperature
falls over time, however, as the waste loses heat to its surroundings. In deeper landfills, this
heat is better retained and degradation occurs faster than in shallower landfills. Water is
generated in the aerobic biodegradation process and required for the anaerobic biodegradation
process. In addition, water movement through a landfill helps to mix the enzymes, bacteria, and
substrate. The subsistence moisture level required by methanogenic bacteria is very low. This
is why gas generation occurs even in the driest of landfills (McBean et al., 1995).
Although moisture content is thought to be an important factor in landfill gas emissions, there is
much variability in the level of emissions from site to site. Typically, emissions in more arid
5-6
DRAFT - DO NOT CITE OR QUOTE
-------
regions are thought to occur over a longer period of time than sites in more temperate climates.
For those sites operated as a wet landfill where leachate has been added or there are other liquid
additions, emissions occur over a much faster rate and there can be a high level of fugitive gas
emissions depending upon how liquid is added to the site.
5.2.3 Gas Emissions
Water and gas flows occur simultaneously in a waste containment facility or contamination
source area as a dynamically-coupled process. As described by Berglund (1998), flow of gas
and water within a landfill can be conceptualized as a trickle bed. The liquid phase trickles over
the waste particles and the gas phase migrates in the remaining pore space. At the present, it is
not possible to effectively integrate all the biochemical reaction and multi-phase transport
mechanisms into one model. Instead, the processes must be uncoupled and discussed
separately. These processes are: (i) percolation of water through a cover system, which was
discussed in Chapter 4; (ii) waste degradation and gas production, which was discussed in
Section 5.2.2; and (iii) gas emissions through a cover system, which is discussed in this section.
Gas flow within and through the cover system of a waste containment facility or remediation
site is mainly pressure driven at gas pressures above about 3kPa, but also responds to
temperature, density, and concentration gradients. Pressures generated by MSW gas may be on
the order of 2.5 to 7.5 kPa for younger landfills located in temperate climates to 0.5 kPa for
older landfills located in arid climates. As gas pressures increase in a waste mass, the gas
travels along the path of least resistance. The final disposition of the generated gas depends on
the engineered controls (e.g., containment systems, gas management system). Gas may be
stored in the waste, migrate through a liner or barrier (if they exist) and into available air space
in the surrounding subsurface, emitted through the cover into the atmosphere, or collected and
treated by a gas management system, if one exists.
Gas emissions may be affected by the gas pressures within a landfill, barometric pressure,
moisture content and gas permeability of the soil components of the cover system, chemical
diffusion rate through a GM barrier component of the cover system, advective flow rate through
any holes in the GM barrier component, and other factors. Barometric pressure is a function of
atmospheric pressure and changes in weather. It responds diurnally to atmospheric tides with a
high in the early morning hours and a low in the afternoon hours. It also responds to changes in
the high and low-pressure systems related to weather conditions. Because gas emissions are
typically pressure driven (i.e., convection rather than diffusion is typically the primary transport
mechanism), gas emissions generally follow the reverse trend of barometric pressures, due to
the effect on pressure gradients. When barometric pressures are highest, gas emissions are
lowest and vice versa. For landfills with active gas collection systems, a change in barometric
pressure should have less impact on gas emissions than occurs for landfills without these
systems.
Even if landfill gases move from the landfill, through the cover system, and into the
atmosphere, some of the gases may be consumed by microbes in cover system soils. The extent
of oxidation is a function of how well the cover is maintained. Often cracks in a cover can
5-7
DRAFT - DO NOT CITE OR QUOTE
-------
result in leaks for the gas to escape to the atmosphere (see Figure 5-2). Data are available
documenting oxidation rates through soil covers. For example, relatively high methane
oxidation rates of 45 g/m2/d were observed in topsoil above a landfill in California (Whalen et
al, 1990); methane concentrations in the air immediately above the topsoil were very low.
Oxidation of methane by soil microbes has been demonstrated in controlled laboratory
experiments (Knightley et al., 1995). The experiments showed that as the flux of methane into
a soil layer decreased, a greater proportion of methane was oxidized. Clearly, however,
widespread documentation of atmospheric methane emissions from soil-covered landfills shows
that microbial methane degradation in cover soils is often not complete. If gas control is needed
at a site, then an active or passive system needs to be installed. The cover material needs to be
maintained to minimize any cracks that will allow a preferential flow path for fugitive
emissions of landfill gas.
Figure 5-3. Surface Cracks at a Landfill
With respect to soils, the pores of the soil must be nearly saturated to prevent gas migration.
Thus, it is not surprising that landfill gas is often detected at the surface of the soil cover
systems. Landfill methane emissions measured at landfill sites and reported in the literature
have ranged from about 0.003 to 3,000 g/m2/d (Bogner and Scott, 1997). In general, the higher
rates were associated with landfills that did not have gas recovery and that were covered with
relatively more permeable and/or drier soils. For example, at the Olinda MSW Landfill in
Southern California, which is covered by a sandy silt soil layer, measured emission rates were
greater than 1,000 g/m2/d prior to installation of a gas collection system. After a gas collection
system was installed, measured gas flux rates were less than 10 g/m2/d. The flux rates were still
lower (less than 0.01 g/m2/d) in the area of the landfill with a gas recovery system and covered
with a clayey silt layer.
5-8
DRAFT - DO NOT CITE OR QUOTE
-------
Characteristics of Selected Gas Emission Models
5.3.1 LandGEM Model for MSW
Gas emission rates for MSW landfills can be difficult to predict. The better the input data, the
better the estimate. Landfill gas emissions vary over time. EPA recommends the use of a first-
order decomposition rate equation to estimate annual emissions over a user-specified time
period. EPA had developed an automated estimation tool for calculating landfill gas emissions.
The is referred to as the Landfill Gas Emissions Model (LandGEM). It uses a Microsoft Excel
interface and is used to estimate emission rates for total landfill gas, methane, carbon dioxide,
nonmethane organic compounds, and individual air pollutants from municipal solid waste
(MSW) landfills (EPA, 2005). The equation that is used in LandGEM is:
where,
QcH4 = annual methane generation in the year of the calculation (m3/year)
i = 1-year time increment
n = (year of the calculation) - (initial year of waste acceptance)
j = 0.1-year time increment
k = methane generation rate (year"1)
L0 = potential methane generation capacity (m3/Mg)
M; = mass of waste accepted in the ith year (Mg)
ty = age of the jth section of waste mass M; accepted in the ith year (decimal years, e.g.,
3.2 years)
LandGEM can use either site-specific data to estimate emissions, or, if no site-specific data are
available, use default parameters. The model contains two sets of default parameters, CAA
defaults and inventory defaults. The CAA defaults are based on federal regulations for MSW
landfills laid out by the Clean Air Act (CAA) and can be used for determining whether a landfill
is subject to the control requirements of these regulations. The inventory defaults are based on
emission factors in EPA's Compilation of Air Pollutant Emission Factors (AP-42) and can be
used to generate emission estimates for use in emission inventories and air permits in the
absence of site-specific test data. (EPA, Chapter 2.4, 1997) The software can be downloaded
from EPA's web site (http://www.epa.gov/ttn/catc/products.html#software). Figure 5.4
provides a screen capture of the user interface.
5-9
DRAFT - DO NOT CITE OR QUOTE
-------
OR.8*
I'S EPA Office of Research nd Development
c
LEAN
A|
T
C
ECHNOLOGY
E N T E
EPA
LandGEM
Landfill Gas Emissions Model
Version 3.02
U.S. Environmental Protection Agency
Office of Research and Development
National Risk Management Re search Laboratory (NRMRL)
and
Clean Air Technology Center (CATC)
Research Triangle Park, North Carolina
May 2005
Figure 5.4 Screen-Capture for EPA's LandGEM Computer Software
The methane generation rate constant, k, is a function of waste moisture content, pH, and
temperature, and nutrient availability to methanogens. In a study by EPA, the value of k for
MSW landfills was estimated to range from 0.003 to 0.21/yr (EPA, 1998) based on field test
data and the results of theoretical models using field test data. For landfills at arid and semi-
arid sites (defined by EPA as sites with less than 640 mm of precipitation per year), EPA's "best
estimate" of k is 0.02/yr (EPA, 1997a). The methane generation potential, LO, is a function of
waste composition. EPA found values for LO ranging from 6.2 to 270 m3/Mg based on
theoretical modeling and field test data for a number of landfills (EPA, 1998). EPA's "best
estimate" of this parameter is 100 m3/Mg (EPA, 1997a) and was obtained from empirical data
from operating landfills using gas extraction data.. Murphy (1998) indicated that the gas
generation rates predicted with the EPA model showed reasonable correspondence to field data
from landfills in arid settings when the default parameters were changed to k = 0.005/yr and L0
= 16 m3/Mg. For most landfills, the EPA parameters can be used with the EPA model to
develop initial estimates of gas generation rates. These parameters can then be adjusted as data
on gas flow rates or emissions are collected over time. Hutric and Soni (1997) describe how
this data fitting may be carried out.
From chemical analysis of landfill gas samples collected at landfills, EPA also developed "best
estimates" of NMOCs and air pollutants for landfills with and without co-disposal of hazardous
waste. EPA's best estimate of the NMOC concentration as hexane is 2,420 ppmv for landfills
that did not have co-disposal of hazardous waste (EPA, 1998). The measured NMOC
concentration reported for the 23 landfill considered by EPA ranged from 240 to 14,300 ppmv
(EPA, 1998). As described by Repa (1994), at the sites in the EPA study, the compounds most
frequently detected in landfill gas included benzene, tetrachloroethene, toluene, trichlorofluoro-
methane, trichloroethane, and vinyl chloride. The compounds detected at the highest average
5-10
DRAFT - DO NOT CITE OR QUOTE
-------
concentration included ethylbenzene, methylene chloride, propane, and xylenes. Effort is
underway to update EPA's landfill gas emission factors. A Cooperative Research and
Development Agreement with the Environmental Research and Education Foundation is
providing cofunding with EPA's Office of Research and Development for conducting field
tests. These data will help in providing more up-to-date landfill gas emission factors. Once
these are released, LandGEM will be updated to included the updated emission factors.
LandGEM does not account for gas storage within the waste mass, nor subsurface gas
migration, which for old unlined landfills can be a significant migration pathway. However, it
is considered a reliable tool in helping to quantify potential gas emissions. If more reliable
estimates are needed, then emission measurements can be conducted using open-path
technology (Modrak et al., 2004 and 2005a and b). Gas extraction rates have been estimated to
be 10 to 60% of the total gas generated (Augenstein et al., 1997). However, collection
efficiency is never precisely known and, if used in the gas generation equation, must be an
assumed value, Hutric and Soni (1997). EPA reports a gas collection efficiency range of 60 to
85%, with an average of 75 percent most commonly assumed. (EPA, Chapter 2.4, 1997)
5.3.2 Diffusion Model for Emissions of Organic Vapors
For hazardous waste landfills and waste piles, if any volatiles are left in the waste when it
reaches the facilities, they are rapidly emitted from the surface of exposed waste. After a cover
system is placed over the waste, emissions of organic vapors occur by diffusion, convection by
barometric pumping, and gas venting.
The model EPA has used for hazardous waste landfills and waste piles assumes diffusion of
volatiles from the waste surface though the cover system (and neglects convective flow due to
changes in barometric pressure) (EPA, 1992). Volatiles in the waste are assumed to be in
equilibrium with air in void spaces of the waste. When the organic vapors reach the surface of
the cover system, they are assumed to be removed by wind (i.e., the constituent concentration at
the cover system surface is assumed to be zero).
5.4 Design of Gas Collection Systems
Gas collection systems are typically designed as part of passive gas management systems or
active gas extraction systems utilizing negative pressure systems. Passive systems are primarily
effective at controlling convective flow (due to pressure and density gradients) and have limited
success controlling diffusive flow. Active systems are effective in controlling both types of
flow. Active systems are preferred when a significant amount of gas is being generated, and
these systems are required for facilities of certain sizes to reduce the amount of gas constituents
released to the atmosphere. Design of gas collection systems can be based on calculated gas
generation rates or vapor emission rates or from the results of field tests (e.g., pump tests).
Some design engineers collect and vent or extract MSW landfill gas with vertical, perforated
collection wells (typically 1 to 3 wells per hectare) without a continuous gas collection layer
beneath the hydraulic barrier component of the cover system. This approach can be justified if
5-11
DRAFT - DO NOT CITE OR QUOTE
-------
the waste itself is sufficiently permeable to gas, if the gas wells are relatively closely spaced, or,
at arid sites, where gas is generated relatively slowly. With gas wells, the gas moves within the
waste to the perforations in the pipe and then flows or is drawn out of the system. Another
approach to venting or extracting gas from a landfill involves installing a continuous gas
collection layer beneath the cover system barrier. With this type of system, shallow gas venting
or extraction pipes will tie into the gas collection layer. Gas collection trenches with periodic
vent or extraction pipes represents a third approach to gas collection beneath the cover system.
Also, a combination of these three gas venting/extraction systems can be used. For active
systems, additional components may include a vacuum blower system, a manifold to connect
multiple wells, off-gas treatment (e.g., enclosed flare, gas-to-energy system, carbon adsorption),
condensate holding tank, and monitoring and control equipment.
In any case (deep wells penetrating the waste, a continuous gas collection layer, beneath the
barrier layer, and/or collection trenches) the system outlets are typically plastic pipes extending
up through the cover system. Gas flow through the pipes can be either passive (vented to the
atmosphere or flared) or active (collected through a header using a blower system to create a
small vacuum). Without a gas management system, gas pressure will build up in the landfill.
Note that with a GM in the cover system and relatively small cover soil thicknesses, gas
pressure can cause GM uplift. Even if the GM is not physically lifted, positive gas pressure
beneath the GM can lower the effective stress at the interface between the GM and underlying
material (e.g., GCL), thereby reducing interface shear strength and potentially contributing to a
slope failure. At several landfill facilities, this latter effect had led to slippage of the GM and
overlying cover materials (Bonaparte et al., 2002) creating high tensile stresses as evidenced by
compression ridges in the cover soil and folding of the GM at the slope toe and tension cracks
in the cover soil near the slope crest.
Based on the above, all of the three types of gas collection systems require careful design
considerations:
. if gas removal is by deep wells, the uppermost pipe perforations should be effective in
capturing gas in the upper layers of waste;
if gas removal is by a gas collection layer beneath the GM and vents, the gas collection
layer should be designed with adequate long-term transmissivity; and
if gas is removed by horizontal collection trenches, some of the trenches should be
placed in close proximity to the bottom of the cover system to prevent gas accumulation
and uplift pressure on the cover system GM.
In general, gas collection systems should be designed with a minimal number of penetrations
through the cover system, as each penetration is a potential location for preferential flow (i.e.,
short-circuiting of gas through the cover system).
For passive systems, a maximum of one well per acre should be included initially (EPA, 1991).
If monitoring of the vents reveals excessively high gas concentrations, then additional wells
can be installed.
In addition to the above, as gases are collected, condensate usually forms because the
5-12
DRAFT - DO NOT CITE OR QUOTE
-------
temperature at the surface is often less than the temperature of the gas. Gas collection systems
often include condensate traps and piping that directs condensate to some collection point (e.g.,
back into a MSW landfill).
5-13
DRAFT - DO NOT CITE OR QUOTE
-------
Chapter 6
Geotechnical Analysis And Design
6.1 Introduction
This chapter provides information on select topics related to cover system geotechnical analysis
and design. The specific topics discussed in this chapter are:
• static slope stability (Section 6.2);
• seismic slope stability and deformation (Section 6.3);
• settlement (Section 6.4);
• steep slopes (Section 6.5); and
• soft waste materials (Section 6.6).
6.2 Static Slope Stability
6.2.1 Overview
Slope stability is a critical issue in the design of cover systems. Slopes on landfills, waste piles,
and other waste containment structures are sometimes quite steep. Sideslope inclinations can
range from flatter than 5H: IV (11.3°) to steeper than 2H: IV (26.6°). For example, cover
systems have been constructed over waste slopes steeper than 1.5H:1V(33.7°) as part of the
remediation of old dumps. This section of the guidance document addresses issues associated
with the static slope stability of cover system components. Both internal and interface
downslope sliding of one or more components are considered. Failure surfaces that extend into
the waste are not addressed herein, but should be considered in slope stability analyses. Special
stability issues associated with cover system sideslopes steeper than about 2.5H: IV and cover
systems installed over soft waste materials are discussed in Sections 6.5 and 6.6, respectively.
The frequency of occurrence of cover system stability problems has been high. More than a
dozen case studies of past problems of this nature are described by Gross et al. (2002) and briefly
discussed in Section 7.4 of this guidance document. One example of a cover system stability
problem is shown in Figure 6-1. The photographs in this figure show a topsoil surface/protection
layer that has slid downslope over a reinforced GCL barrier in a cover system that did not
contain an internal drainage layer. Figure 6-2 shows another example, this one involving a
topsoil surface/protection layer and underlying sand drainage layer that has slid over a textured
HDPE GM barrier. In this case, the sand drainage layer (specified hydraulic conductivity of
1 x 10"5 m/s) had inadequate flow capacity and the drainage layer outlets were constricted. With
GMs, GCLs, CCLs, GTs, and GCs commonly used in a variety of cover system configurations,
the stability of potential low shear strength materials and interfaces must be considered for most
DRAFT - DO NOT CITE OR QUOTE
6-1
-------
Figure 6-1. Example of Cover System Slope Stability Problem. The Topsoil
Surface/Protection Layer Slid Downslope Over the Reinforced GCL Barrier.
DRAFT - DO NOT CITE OR QUOTE
-------
Figure 6-2. Example of Cover System Slope Stability Problem. The Topsoil Surface/
Protection Layer and Underlying Sand Drainage Layer Slid Downslope Over
the Textured HOPE GM Barrier.
designs. Significantly, past failures have involved sliding along each of the geosynthetic
interfaces listed in Table 6-1.
Table 6.1 Interfaces upon which cover system components have undergone sliding.
• Topsoil surface/protection layer sliding on:
GT
GM
GCL
CCL
• Sand drainage layer sliding on:
GT
GM
GCL
CCL
GN drainage layer sliding on GM
GC drainage layer sliding on GM
GT sliding on GM
GM sliding on:
GT
GCL
CCL
GCL sliding on:
CCL
prepared subgrade
6.2.2 Limit Equilibrium Analyses
6.2.2.1 Overview
The simplest limit equilibrium (LE) formulation to analyze the slope stability of cover systems
assumes infinite slope conditions and neglects the stabilizing influences of passive soil resisting
DRAFT - DO NOT CITE OR QUOTE
6-3
-------
forces at the toe of the slope, any true cohesion/adhesion in cover system materials and
interfaces, and tension in the geosynthetic layers. More sophisticated LE formulations account
for these factors. Both the infinite slope and more sophisticated LE formulations are discussed
below. In all of the closed-form, two-dimensional LE solutions, force equilibrium is satisfied in
the directions normal and parallel to the slope, but moment equilibrium is ignored.
6.2.2.2 Infinite Slope
For cover system geometries where the cover soil thickness is constant, infinite slope equations
provide a simple and conservative basis for design. Equations can be formulated in terms of: (i)
total unit weights of the cover system materials and boundary water pressures; or (ii) buoyant
unit weights and body seepage (or drag) forces. In keeping with the approach of Giroud et al.
(1995a), equations are formulated herein using buoyant unit weights and seepage forces.
Body seepage forces occur in cover systems when water infiltrating the cover system develops a
significant flow component in the downslope as opposed to vertical downward direction. This
occurs, for example, when infiltration is blocked by a hydraulic barrier. If the rate of infiltration
is sufficient, hydraulic head will build up above the barrier layer and induce downslope flow.
Downslope flow of water has a destabilizing effect on the cover system. The seepage force per
unit volume on soil particles in the direction of laminar flow is expressed as:
fw=Ywi (Eq.6.1)
where: fw = seepage force per unit volume (N/m3); yw = unit weight of water (N/m3); and i =
hydraulic gradient (dimensionless). The concept of a seepage force, Fw (N) (acting parallel to the
slope), and buoyant unit weight, Wb (N) (acting vertically), in an infinite soil slope underlain by
a hydraulic barrier is illustrated in Figure 6-3. For a 1-m thick cover system at a 3H: IV (18.4°)
slope and with water flowing in the entire soil thickness, the water induces a downslope body
seepage force of 3 kPa.
If there is no water flow in an infinite slope, the slope stability factor of safety is given by
(Giroud et al., 1995a):
tanp yttsinp
where: FS = factor of safety (dimensionless); ((); = angle of internal or interface friction for the
critical potential slip surface (degrees); a; = adhesion (for an interface) or cohesion (for internal
strength) for the critical potential slip surface (Pa); |3 = slope angle (degrees); yt = total unit
weight of material above the critical potential slip surface (N/m3); and t = thickness of material
above the critical potential slip surface (m). Use of this equation assumes that there is a unique
critical potential slip surface in the cover system. For the case of no adhesion or cohesion (a; =
0), Eq. 6.2 reduces to the classical solution:
FS = tan <(); /tanp (Eq. 6.3)
For a hydraulic barrier system, two conditions need to be considered: (i) stability above the
hydraulic barrier; and (ii) stability below the hydraulic barrier. These two conditions must be
DRAFT - DO NOT CITE OR QUOTE
6-4
-------
considered because effective stresses above and below a non-porous hydraulic barrier, such as a
GM, are different. The infinite slope factor of safety for "full flow" (tw = thickness of water flow
parallel to the slope = t in Figure 6-3) parallel to the slope along an internal or interface slip
surface above the hydraulic barrier is (Giroud et al., 1995a):
FS -
rsA -
ysattsinp
(Eq. 6.4)
where: FSA = factor of safety for critical potential slip surface above the hydraulic barrier
(dimensionless); §a = angle of internal or interface friction for the critical potential slip surface
above the hydraulic barrier (degrees); aa = cohesion (for internal strength) or adhesion (for an
interface) for the critical potential slip surface above the hydraulic barrier (Pa); yb = average
buoyant unit weight of material above the critical potential slip surface (N/m3); and ysat = average
saturated unit weight of material above the critical potential slip surface (N/m3); and all other
terms are as defined previously. The buoyant unit weight, yb, is equal to total unit weight, yt,
minus the unit weight of water, yw.
Hydraulic Barrier
Figure 6-3. Seepage Force and Buoyant Unit Weight for a Soil Layer Overlying a
Hydraulic Barrier on an Infinite Slope (modified from Giroud et al, 1995a).
This factor of safety can be compared with the factor of safety expressed by Eq. 6.2 for the case
of no water flow. The comparison shows that for typical soils:
• FSA MI HOW /FSno flow -0.5 if a; = 0; and
• FSAfullflow/ FSno flow ~0.9 if (|)i = 0.
Based on these results, for slip surfaces located above the hydraulic barrier, the factor of safety
can decrease by a factor of two due to water flow parallel to the slope if shearing resistance is
generated primarily through friction.
DRAFT - DO NOT CITE OR QUOTE
6-5
-------
The factor of safety ratios presented above are based on the assumption that the shear strength
properties, (j)a and aa, are not influenced by the presence of water. If the presence of water
reduces the magnitudes of these parameters, the effects noted in the above comparison would be
even more substantial.
The infinite slope factor of safety for "full flow" parallel to the slope along an internal or
interface slip surface below a non-porous hydraulic barrier is given by (Giroud et al., 1995a):
FS
tanp Y
where: FSB = factor of safety for critical potential slip surface below the hydraulic barrier
(dimensionless); fa and ab are the internal or interface shear strength parameters for the critical
potential slip surface below the hydraulic barrier; and all other terms are as defined previously.
It should be noted that the shear strength parameters §\, and ab, used in Eq. 6.5, will typically be
different than the parameters (j)a and aa, used in Eq. 6.4, as the interfaces in the two equations are
different. This factor of safety is to be compared with the factor of safety expressed by Eq. 6.2
for the case of no water flow. The comparison shows that for typical soils:
• F SB full HOW / F Sno flow = 1 if a; = 0; and
• FSAfuiiflow/FSnoflow~0.9if(t)i = 0.
Based on these results, the factor of safety along critical potential slip surfaces below the
hydraulic barrier is only affected to a relatively minor degree by water flow above the hydraulic
barrier.
The final infinite slope case to be considered is for "partial-depth" flow (tw < t in Figure 6-3)
parallel to the slope. The appropriate equations are (Giroud et al., 1995a):
(Eq. 6.6)
tw) + ysattw tanp Yt(t-tw)
and
FSB =- + - - (Eq. 6.7)
tanp yt(t-tw
where: tw = thickness of water flow parallel to the slope (m), as defined in Figure 6-3, and all
other terms are as defined previously.
Based on the foregoing equations, the effect of water flow on the stability of a cover system is
much greater if the slip surface is above the hydraulic barrier than if it is below the hydraulic
barrier. The reasons for this can be summarized as follows:
DRAFT - DO NOT CITE OR QUOTE
6-6
-------
• The main effect of water flowing downslope within a cover system slope is the
significant decrease in the effective normal stress above the hydraulic barrier.
• Other effects of water flowing downslope within a cover system are a slight increase in
the effective normal stress below the hydraulic barrier layer and a slight increase in the
shear stress above and below the hydraulic barrier.
• As a result of the changes in effective normal stress, the frictional component of shear
strength decreases significantly above the hydraulic barrier but decreases only slightly
below the hydraulic barrier.
• As a result of the changes in shear strength and the slight increase in shear stress, the
factor of safety is significantly affected above the hydraulic barrier and only mildly
affected below the hydraulic barrier.
It can also be inferred from the above assessment that waste-generated gases beneath a cover
system effect the stability of the interface between a non-porous hydraulic barrier and an
underlying material by decreasing the frictional component of shear strength along the interface
while the shear stress along the interface remains unchanged. This is one reason why gases may
need to be collected and controlled via a gas collection layer, gas wells, or other means. One
example of a cover system stability problem caused by gas pressures is described in Section 7.7.
Briefly, gas generated in a MSW landfill uplifted the GM barrier of a cover system and resulted
in the GM and overlying materials moving downslope over a GT. Though the landfill had
vertical gas extraction wells, the upper portions of the wells were not perforated. As a
consequence, gas accumulated beneath the cover system, generating uplift pressures on the
underside of the GM.
6.2.2.3 Slope of Finite Length
Equations for the LE evaluation of sloping geosynthetic-soil layered systems (such as a cover
system) for a slope of finite length have been presented by Giroud and Beech (1989), EPA
(1991), Koerner and Hwu (1991), McKelvey and Deutsch (1991), Bourdeau et al. (1993),
Druschel and Underwood (1993), Giroud et al. (1995a,b), Soong and Koerner (1997), Koerner
and Daniel (1997), and Koerner and Soong (1998), among others. The most detailed treatments
of the subject have been presented by Koerner and coworkers and Giroud et al. (1995a,b).
Giroud et al. (1995b) have shown that compared with the method they present, the method
utilized by Koerner and coworkers is more rigorous, but somewhat more complicated to use
because it requires solution of quadratic equations. The formulation by Giroud et al. (1995a,b)
involves an approximation that allows expression of the factor of safety as a closed-form
algebraic equation where each term in the equation has a distinct physical meaning and is
sufficiently accurate for practical purposes. The simpler formulation is presented below, but
either method is acceptable when properly applied.
The two-part wedge considered by Giroud et al. (1995a,b) is illustrated in Figure 6-4. For this
condition, the slope stability factor of safety for a slope with constant soil thickness above the
critical potential slip surface and for the case of no water flow (tw = 0 in Figure 6-4) is given by:
DRAFT - DO NOT CITE OR QUOTE
6-7
-------
T/h
tanp yt t h ivsin(2p)cosi
Ytt
(Eq. 6.8)
where: (|)s = angle of internal friction for the soil material (i.e., protection layer and/or granular
drainage layer) above the critical potential slip surface (degrees); cs = cohesion of soil material
above the critical potential slip surface (Pa); h = height of slope (m), as defined in Figure 6-4; T
= geosynthetic tension above the potential slip surface (N/m); and all other terms are as defined
previously.
Wedge 2
Wedge 1
C
Hydraulic Barrier
Figure 6-4. Definition of Two-Part Wedge and Flow Thickness for the Case of a Slope of
Finite Height (modified from Giroud et al., 1995a).
Eq. 6.8 consists of five terms, each of which has physical significance. The significance of each
term is as follows:
• The first term quantifies the contribution of the frictional component of the critical
interface or internal shear strength to stability (i.e., the frictional component along line
segment AB in Figure 6-4).
• The second term quantifies the contribution of the adhesion component of the critical
interface or internal shear strength to stability (i.e., the adhesion component along line
segment AB in Figure 6-4).
• The third and fourth terms quantify the contribution of the toe buttressing effect, which
results from the shear strength of the soil located at the toe of the slope above the slip
surface (i.e., the soil shear strength along line segment BC in Figure 6-4). Both terms
depend on the soil internal friction angle, whereas only the fourth term depends on the
soil cohesion.
• The fifth term quantifies the contribution to the factor of safety of any tension in the
geosynthetics located above the slip surface (which may include one or more
DRAFT - DO NOT CITE OR QUOTE
-------
geosynthetics specifically used as reinforcement).
The case of partial-depth and full-depth flow of water for a slope of finite height was addressed
by Giroud et al. (1995a). For the case of a slope of uniform thickness above the critical potential
slip surface, the factor of safety above the hydraulic barrier may be calculated using the
following equation:
yt(t-tw) + ybtw tan^a | aa/sinp
yt(t-tw) + ysattw
L^ . J (Eq. 6.9)
ct/h Y cos* T/h
lyt(t-tw) + ysattwXsinpcos(p + k)J yt(t-tw) + ysattw
where: t w = thickness of water in Wedge 1 (m), as defined in Figure 6-4; and all other terms are
as defined previously. For potential slip surfaces below a non-porous hydraulic barrier:
F§ =tan^b| ab/sinp ; f Ytft-Q + YbC Y* Y sin^ 1
tanp yt(t-tw) + ysattw 1 yt(t-tw) + y.attw i h 1 sin(2B)cos(B + (|),) J
i i L v w s i sdL w y i L v w s i aaL wyvxy \ i / \i isxy /T"" s~ 1 /~\\
f ., . (Eq. 6.10)
[ cst/h Y cosk ] T/h
When there is full flow of water in Wedge 1 (t*w = t) as well as in Wedge 2 (tw = t), Eq. 6.9 gives
the following equation for the factor of safety for a critical potential slip surface above the
hydraulic barrier:
ysattsinp y^ s .
(Eq. 6.11)
s | T/h
and Eq. 6.10 reduces to:
FSB =
tanp ysattsinp ysat s S
(Eq. 6.12)
s | T/h
where all terms are as defined previously.
DRAFT - DO NOT CITE OR QUOTE
6-9
-------
Another case sometimes encountered is that of a tapered cover soil thickness, as illustrated in
Figure 6-5. For this geometry, the factor of safety for the case of no water flow is given by the
equation:
FS =
tanc^ tt
tanp tm
+±-(-
tv
sinc|)s
yttb sinp h ^sin(2p)cos(P
cos<|)s ^ T/h
ythl sinpcos(P + (|)s) J yttt
(Eq. 6.13)
where:
tavs=(ta+tb)/2
(Eq. 6.14)
tavg = average thickness of soil layer between points A and B, which are defined in Figure 6-5
(m); ta = thickness of soil layer at point A (m), as defined in Figure 6-5; tb = thickness of soil
layer at point B (m), as defined in Figure 6-5; and all other terms are as defined previously.
ltoe
Hydraulic Barrier
Figure 6-5. Definition of Slope with a Tapered Soil Layer (from Giroud et al., 1995b).
Eq. 6.13 can also be used to calculate the factor of safety for a partly tapered slope of height h,
illustrated in Figure 6-6, by calculating an average soil thickness for the entire slope using the
equation:
U^fl + Vl + T-fl-Vl (Eq.6.15)
where: hu = height of slope above the slope grade break (m), as illustrated in Figure 6-6, and all
other terms are as defined previously.
DRAFT - DO NOT CITE OR QUOTE
6-10
-------
Hydraulic Barrier
Figure 6-6. Definition of Slope with a Partly Tapered Soil Layer (from Giroud et al.,
1995b).
The equations presented above provide closed-form solutions to a variety of cover system slope
stability situations. Some situations are too complex, however, to address using closed-form
solutions and are more easily evaluated using commercially available two-dimensional slope
stability computer software (e.g., PCSTABL5M (Achilleous, 1988), UTEXAS4 (Wright, 1999),
XSTABL (Sharma, 1994), or SLOPE/W (available from Geo-Slope Int. Ltd., Alberta, Canada)).
Available software has the advantage over closed-form solutions in that it can be applied to non-
uniform slope, soil cover, and hydraulic head conditions, and can incorporate a pseudo-static
seismic coefficient for use in seismic stability evaluations.
It is noted that the above two-dimensional LE methods are based on a plane-strain condition and
do not consider the shear resistance along the two sides of the slide mass that parallel the
direction of movement. A two-dimensional analysis, however, is considered appropriate for
cover system design because it yields a conservative estimate of the slope stability factor of
safely for design geometries encountered in cover systems. The degree of conservatism
decreases as the cover system geometry approaches a two-dimensional configuration (i.e., the
ratio of cover system slope width to slope length increases). For the majority of cover system
geometries, the incremental increase in stability calculated by considering three-dimensional
effects will be negligibly small.
The LE method is useful for evaluating cover system stability under most conditions but is
subject to several limitations. With the LE method, material and interface shearing resistances
are assumed to be independent of displacement. For geosynthetic materials and interfaces,
however, mobilized shearing resistance is not constant but increases with increasing
displacement to a peak value. For many materials and interfaces, the shear resistance decreases
with increasing displacement after reaching the peak, and ultimately reaches a "residual" value
(Figure 6-7). This behavior is sometimes referred to as "strain-softening." In using the LE
method, judgment must be applied to the selection of shear strength values for strain-softening
materials (i.e., peak, residual, or some other value). The LE method is similarly limited with
respect to tension forces in cover system geosynthetic components and, therefore, cannot be used
to estimate the magnitude and distribution of stresses and deformations in these components.
These limitations of the LE method can be overcome by using another slope stability evaluation
method, stress-deformation analyses discussed below in Section 6.2.3.
DRAFT - DO NOT CITE OR QUOTE
6-11
-------
0 5 10 15 20 25 30 35 40 45 50 55 60
Shear Displacement (mm)
Figure 6-7. Results of Direct Shear Test on a GCL Illustrating Peak and Large-
Displacement Shearing Resistances at Different Normal Stresses (an).
As a final comment on the LE equations presented in this section, the equations incorporate
terms to account for material internal cohesion or interface adhesion and for geosynthetic
tension. Caution should be taken when selecting cohesion or adhesion and geosynthetic tension
values for design. As suggested by Koerner and Daniel (1997), cohesion and adhesion values
should be used only when there is clear physical justification. From the analysis results
presented previously, characterization of internal or interface shearing resistance by a cohesion
or adhesion term instead of a friction angle will greatly affect the results of slope stability
analyses when hydraulic heads are present in the cover system. In general, geosynthetic tension
should not be in the equations unless the design includes a geosynthetic reinforcement layer.
Other types of geosynthetics, such as GMs, GNs, GCs, etc., are not designed to permanently
transmit tensile loads, are potentially subject to significant tensile creep, and typically have a low
tensile modulus (which means that the geosynthetic must elongate to generate tension). Even if
geosynthetic reinforcement is used, it should only be relied upon for the tensile force that it can
generate at a specified acceptable level of deformation. This acceptable level of deformation
must be selected considering the overall performance of all system components.
6.2.3 Stress-Deformation Analyses
Stress-deformation analysis methods may be used for cover system design when the limitations
of LE methods are considered significant. The primary advantage of stress-deformation methods
is their ability to account for the stress-strain response of materials and interfaces and, therefore,
to predict the distribution of stresses and strains within the cover system components,
particularly geosynthetic components. Stress-deformation methods can also account for the
effects of construction sequencing. The primary disadvantage of stress-deformation methods is
the relatively large effort required to obtain material stress-deformation relationships and
perform the calculations compared to the effort required with LE methods.
DRAFT - DO NOT CITE OR QUOTE
6-12
-------
Several studies have been published on the application of stress-deformation methods to cover
systems. For example, Long et al. (1993, 1994) described a finite difference model
(GEOSTRES) that considers stress equilibrium and strain compatibility. GEOSTRES uses
inelastic, non-linear springs to model the shear resistance-displacement behavior at each
interface and to model the axial load-displacement behavior within each component. Wilson-
Fahmy and Koerner (1992, 1993) adopted a two-dimensional finite element model to account for
stress equilibrium and deformation compatibility in stability analyses of soil-geosynthetic
systems on slopes.
6.2.4 Shear Strength Parameters
It is recommended that laboratory testing using project-specific materials, coupled with testing
procedures and conditions representative of the anticipated field application, be performed to
establish design shear strength parameters on a project by project basis. Sabatini et al. (2001)
have shown that for a given factor of safety, designs based on project-specific laboratory testing
programs are more reliable and less prone to slope instability than designs that utilize shear
strength parameters obtained from more general sources, such as databases or the published
technical literature.
The various methods used for laboratory shear strength testing of soils are well known and are
fully described in a number of geotechnical textbooks and laboratory guides (Lambe, 1951;
Holtz and Kovacs, 1981; Bardet, 1997). The most commonly used methods for laboratory shear
strength testing of soils are the triaxial compression test and direct shear test.
Currently there are several types of laboratory devices available for the evaluation of shear
strength of geosynthetic materials and interfaces. These laboratory devices include:
• large-scale (300 mm x 300 mm) direct shear box specified by ASTM D 5321;
• conventional (50 to 100 mm square or circular) direct shear box with testing generally
following ASTM D 3080;
• torsional shear device (ASTM standard under development);
• tilt table; and
• large-displacement shear box.
A summary of the advantages and disadvantages of the first four devices is presented in Table
6-2. Shallenberger and Filz (1996) described the capabilities and limitations of the large-
displacement shear box. More recently, Marr (2001) discussed the attributes of test equipment
and methods used to evaluate the shear strength of geosynthetic materials and interfaces.
Most project-specific laboratory testing being performed presently uses the ASTM standard 300
mm x 300 mm direct shear box. The large scale of this box is advantageous due to the structure
of many geosynthetics, which requires a large test specimen to achieve a representative size of
material for testing.
DRAFT - DO NOT CITE OR QUOTE
6-13
-------
Table 6-2. Summary of advantages and disadvantages associated with test devices for
measuring interface shear strength (modified from Gilbert et al., 1995).
Test Device
Large-scale direct shear box
Conventional direct shear box
Torsional shear device
Tilt table
Advantages
Industry standard
Large scale
Large displacement
Minimal boundary effects
Large experience base with soil
Large normal stress
Inexpensive
Unlimited continuous displacement
Minimal machine effects
Minimal boundary effects
Inexpensive
Disadvantages
Machine friction
Load eccentricity
Limited continuous
displacement
Limited normal stress
Expensive
Machine friction
Load eccentricity
Small scale
Limited displacement
Boundary effects
Machine friction
Anisotropic shearing
Small scale
Expensive
Small experience base
Limited continuous
displacement
Limited normal stress
No post-peak behavior
Large effort to prepare sample
Project-specific shear strength testing programs are designed to simulate the anticipated field
conditions by selecting appropriate testing procedures and conditions. These include the soil
compaction conditions (i.e., water content and density), soil consolidation stress and time,
wetting conditions for the materials and interfaces, range of applied normal stresses, direction of
shear for geosynthetic interfaces, and shear displacement rate and magnitude. The potential
effects of many of these testing conditions on measured interface shear strength parameters are
reported in the literature (e.g., Martin et al., 1984; Saxena and Wong, 1984; Bonaparte et al.,
1985; Williams and Houlihan, 1987; Seed et al., 1988; Giroud et al, 1990; Seed and Boulanger,
1991; Swan et al., 1991; Pasqualini et al., 1993; Stark and Poeppel, 1994; Bemben and Schulze,
1995; Gilbert et al., 1995; Nataraj et al., 1995; Bonaparte et al., 1996; Gilbert et al., 1996;
Shallenberger and Filz, 1996; Stark and Bid, 1996; Stark et al., 1996; Dove et al., 1997; Bid and
Stark, 1997; Sharma et al., 1997; Daniel et al. 1998; De and Zimme, 1998; Fox et al., 1998;
Sabatini et al. 1998; Snow et al., 1998; Li and Gilbert, 1999; Breitenbach and Swan, 1999).
Particular attention should be given to the following:
• Testing should be performed with materials and boundary conditions representative of
the anticipated field conditions.
• Soils used in the tests should be compacted to representative field conditions. The
compaction moisture content for CCLs used in a direct shear testing program should be
near the upper limit of acceptable moisture content and near the lower limit of dry unit
weight allowed by the construction specification.
• For GM/CCL interface shear tests, a variety of opinions exist with regard to the
application of additional moisture to the interface just prior to assembly of the test
DRAFT - DO NOT CITE OR QUOTE
6-14
-------
specimen. Options include not adding moisture, lightly or moderately "spritzing" water
onto the CCL, or submerging the assembled sample. The rationale for any of these
techniques is to simulate suspected installation (e.g., rainfall or moisture conditioning) or
post-installation (e.g., condensation collecting at the interface or consolidation-induced
water movement to the interface) increases in CCL moisture content at the interface.
Counterbalancing these potential mechanisms for moisture content increase at the
interface is the effect of thermal gradients typically induced in CCLs beneath GMs prior
to covering the GMs with soil. The thermal gradients tends to induce water vapor
migration away from the hotter interface and into the underlying cooler soil (Bowders et
al., 1997a). Another factor to consider is the post-compaction thixotropic effect
identified by Shallenberger and Filz (1996), wherein residual interface shear strengths
were found to increase with "curing time" after sample preparation. Given all of these
factors, the design engineer must give careful consideration as to the application of
additional moisture to the interface just prior to assembly of the test specimen.
• Hydration (soaking) times for GCL samples should be adequate to achieve minimum
strength. Daniel and Scranton (1996) showed that hydration times of 24 hours were
sufficient for small, 64-mm diameter samples. Koerner and Daniel (1997) noted,
however, that complete hydration of relatively large (300 mm x 300 mm) direct shear
tests samples takes longer than traditionally required for hydration of soils in relatively
small direct shear boxes. Gilbert et al. (1996) reported hydration times, as determined by
cessation of GCL swelling under constant normal stress, for reinforced GCLs of up to 25
days. However, Gilbert et al. (1996) used deionized water as the permeating liquid
(which increases swell potential), and Daniel et al. (1993) showed that full hydration is
not necessary to achieve minimum shear strength. Given this information, an acceptable
approach to GCL hydration is to monitor vertical deformation of the GCL and continue to
hydrate until these deformations have ceased under the applied normal stress (see
discussion of normal stress below). When this procedure cannot be performed, a
minimum hydration time of 72 hours is recommended for GCLs to be tested in a 300 mm
x 300 mm direct shear box. It should be remembered that without adequate hydration
time, the measured GCL strength may be larger than the fully hydrated strength.
• Testing conditions must adequately reflect the field consolidation conditions of the GCL
or CCL components. GCLs hydrated as indicated above will be fully consolidated under
the normal stress applied during hydration. Consolidation requirements for CCLs may be
established using ASTM D 3080. Specimen consolidation times of 48 hours or more
may be required for some CCL materials. For both GCLs and CCLs, the normal stress
applied during hydration should be equal to the normal stress applied by the cover system
in the field, if the full thickness of overlying cover materials is to be placed quickly.
Alternatively, a more conservative approach would be to apply a normal stress during
hydration equal to only a portion of the overburden stress (e.g., one-third or one-half) that
will exist once the cover system is fully constructed. In this latter approach, after
hydrating the GCLs, they should be consolidated at the normal stress associated with the
full weight of the overlying cover system layers. Under the low normal stresses
associated with most cover systems, GCLs will typically swell during hydration.
• ASTM D 5321 and ASTM D 6243 recommend that tests be performed at a minimum of
three normal stresses, with each test conducted on a new test specimen. The three
DRAFT - DO NOT CITE OR QUOTE
6-15
-------
selected normal stresses should bracket the normal stress applied by the cover system to
the material or interface being tested. This is important because many of the materials
used in cover systems exhibit a non-linear relationship between internal or interface shear
resistance and normal stress. For cover systems, the applicable range of normal stresses
will typically be in the range of about 5 to 40 kPa. Uniformity of normal stress over the
entire test specimen must be maintained during hydration, consolidation and shearing so
as to avoid stress concentrations.
• Shear displacement rates should be selected considering the type of slope stability
analysis to be performed and the types of potentially critical materials or interfaces to be
tested. For geosynthetic/geosynthetic interfaces (excluding GCLs), the maximum rate
allowed by ASTM D 5321 of 0.08 mm/s will generally be acceptable. For long-term
stability conditions where the potentially-critical material or interface includes a CCL or
GCL component, the shear displacement rate should be as slow as reasonably achievable;
the default shear displacement rate of 0.017 mm/s given in ASTM D 5317 is too fast to
achieve drained shearing conditions for CCLs and GCLs. Procedures for estimating
shear rates to obtain fully-drained conditions for CCLs are given in ASTM D 3080.
Procedures and data for estimating shear rates to obtain fully-drained conditions for
GCLs are given in ASTM D 6243. It is noted, however, that it may not be necessary to
achieve fully-drained test conditions to obtain test results suitable for long-term analyses.
Available data suggest that for design purposes, a shear displacement rate of not more
than 0.0005 mm/s will produce test results appropriate for use in slope stability analyses
involving GCL materials and interfaces. In contrast, for the evaluation of seismic
stability, shear displacement rates should be as fast as reasonably achievable. For both
conditions, testing should be performed using samples fully-consolidated under the
applied normal stresses.
• Tests should be carried out to a shear deformation adequate to evaluate both the peak and
large-displacement shear resistance of the material or interface being tested. Many
geosynthetic/geosynthetic and soil/geosynthetic interfaces exhibit very significant post-
peak reductions in shear strength (Figure 6-7). ASTM D 5321 states that one should "run
the test until the applied shear force remains constant with increasing displacement." To
achieve a large-displacement shear condition (defined as a relatively-flat, post-peak shear
stress versus shear displacement line) in a direct shear test, shear displacements of 50 mm
or more may be necessary. It should also be noted that this large-displacement shear
strength is close to, but typically not as low as, the absolute minimum (i.e., residual)
shear strength of the material or interface. Residual shear resistances may not occur until
shear displacements reach 200 mm, or more. Torsional ring shear testing (Stark and
Poeppel, 1994) can be used to evaluate residual shear strengths for soils and
geosynthetics for which representative samples can be produced for the small size and
torsional shearing mode of this type of test. Alternatively, large-displacement shear box
testing (Shallenberger and Filz,1996) can be used to evaluate residual shear strengths for
larger-size test specimens in a linear displacement mode. For most practical design
applications, true residual strength can be estimated to an acceptable degree of accuracy
as 90 to 95% of the large-displacement strength obtained from a 300 mm x 300 mm
direct shear test.
• Multi-component cover systems may have more than one potentially-critical slip surface.
DRAFT - DO NOT CITE OR QUOTE
6-16
-------
The shear test program for a project may need to consider several materials or interfaces.
• Some materials exhibit significant manufacturing variability. For example, the degree of
texturing on GMs and the amount of internal-reinforcing in needlepunched GCLs has
been observed to vary significantly from lot to lot. This variability should be considered
both in design and in the selection of project QC/QA protocols.
• Test results can be interpreted in terms of a secant friction angle that varies as a function
of normal stress, or by a tangent friction angle and apparent cohesion (or adhesion for
interface strength) applicable to the range of considered normal stresses. Both of these
approaches are illustrated in Figure 6-8. For cover system applications, internal and
interface shear strength parameters should be defined in terms of a secant friction angle
for cases where hydraulic heads could develop in the cover soil. Since the apparent
cohesion or adhesion may not be a true material or interface property, the use of this
parameter with high heads (relative to the total normal stress) could lead to an over-
prediction of the true slope stability factor of safety. Also, as previously mentioned,
Koerner and Daniel (1997) suggested that cohesion and adhesion values should be used
only when there is clear physical justification.
All of the foregoing factors should be considered in designing a laboratory shear testing program
to evaluate internal and interface shear strengths and in using the results of the program in slope
stability analyses.
Several other factors may affect long-term shear strength properties of the cover system
materials and interfaces. For example, in cold regions, freeze-thaw may reduce the shear
strength of cover system CCLs and CCL/geosynthetic interfaces. Research has shown that many
CCLs undergo significant change in soil fabric and reduction in shear strength as a result of
freeze-thaw cycling (e.g., Nagasawa and Umeda, 1985; Othman et al., 1994). A case study
illustrating how this problem contributed to slope failure for a landfill cover system in Ohio is
presented in Section 7.4.3. In addition, both heating and cooling result in soil moisture
migration, which can cause changes in material and interface, shear strengths (e.g., Daniel et al.,
1993). Furthermore, long-term creep may also be significant, particularly in geosynthetic
components. No consistent standard of practice presently exists for directly addressing the
potential effects of all of these factors on cover system stability. These factors may be indirectly
accounted for through the use of higher minimum acceptable factors of safety, when appropriate,
or through placement of a greater thickness of cover soil above the critical layers for thermal
insulation and isolation from environmental factors.
DRAFT - DO NOT CITE OR QUOTE
6-17
-------
Shear Displacement (Q)
(a)
Ytj = Tangent Friction Angle
3aj = Apparent Adhesion/Cohesion
Normal Stress (On)
(b)
(|)sj = Secant Friction Angle
Normal Stress
(C)
Figure 6-8. Interpretation of Cover System Interface or Internal Shear Test: (a) Test
Results for Peak Strengths; (b) Tangent Friction Angle, $t\, and Apparent
Adhesion or Cohesion, aai; and (c) Secant Friction Angle, $s\- Similar
Interpretations are Applied to Large-Displacement and Residual Conditions.
DRAFT - DO NOT CITE OR QUOTE
6-18
-------
6.2.5 Construction Considerations
The placement of soil over a slope with underlying low shear strength materials or interfaces will
induce shear stresses that can reduce slope stability. These shear stresses result from the
operation of construction equipment on the slope, the weight of the soil, and, if the soil is pushed
down the slope, from the moving soil itself. Construction-induced stresses have been
investigated by McKelvey and Deutsch (1991) and Koerner and Daniel (1997). These references
present closed-form LE equations that can be used to evaluate the effect of construction
equipment operation on cover system stability. The clear recommendation that comes out of
these investigations is that cover soils should be placed over low shear strength materials and
interfaces from the bottom of the slope upward and not from the top of the slope downward
(Figure 6-9).
The following comments are provided with respect to placement of soil materials in cover
systems:
• By placing cover soils from the bottom of the slope upward, a passive, stabilizing soil
wedge is established at the toe of slope prior to placement of soil higher on the slope.
The operation of construction equipment over this lower wedge tends to compact and
strengthen the wedge.
• Relatively small, wide-track dozers (i.e. low-ground pressure dozers) are recommended
for placing the soil cover material. This type of equipment limits both the dynamic force
imparted to the slope during acceleration and braking and the tractive force applied
through the dozer tracks.
• Downslope dynamic forces can be limited further by limiting the dozer speed on the
slope and by instructing the dozer operator to avoid hard breaking, particularly when
backing downslope.
By application of the construction procedures described above, construction-induced impacts to
the stability of a cover system slope (designed to conventional slope stability factors of safety
described next in Section 6.2.6) are minor. For other conditions (e.g., lower factors of safety
than recommended in Section 6.2.6, placement of soil from the top of slope downward, use of
large construction equipment) construction-stage stability should be checked using the
procedures described by McKelvey and Deutsch (1991) or Koerner and Daniel (1997).
6.2.6 Factors of Safety
LE analysis methods provide a calculated slope stability factor of safety (FS). Minimum
acceptable FS values for cover systems depend on project-specific conditions and uncertainties.
For example, when cover systems include strain-softening materials or interfaces, differing
minimum factors of safety are often applied to peak strength analyses and analyses based on
large-displacement or residual strength. Other criteria may also influence selection of a
minimum acceptable FS, including regulatory requirements, reliability of laboratory test
methods, similarity between laboratory testing conditions and field conditions, completeness of
laboratory test data, uncertainty with respect to other design input parameters (e.g., unit weights,
hydraulic heads, geometry), and consequences of slope failure.
DRAFT - DO NOT CITE OR QUOTE
6-19
-------
Crest
Anchorage-
Hydraulic Barrier
Toe
Buttress
(a)
Hydraulic Barrier
(b)
Figure 6-9. Cover Soils Should be Placed Over Low Shear Strength Materials and
Interfaces from the Bottom of the Slope Upward (a) and not from the Top of
the Slope Downward (b).
Previous agency guidance on selecting slope stability factors of safety was given in EPA (1988).
The FS values given in the 1988 document were meant to apply to excavation and embankment
(soil) slopes used in the construction of landfills and surface impoundments. As the reported FS
values represent general guidance, however, they have sometimes been cited as criteria for the
design of cover systems. The values from EPA (1988) are given in Table 6-3 below.
DRAFT - DO NOT CITE OR QUOTE
6-20
-------
Table 6-3. Previously-recommended minimum FS values (modified from EPA, 1988).
Consequences of Slope Failure
No imminent danger to human life or major
environmental impact if slope fails
Imminent danger to human life or major
environmental impact if slope fails
Uncertainty of Strength
Small1
1.25
1.5
Measurements
Large2
1.5
2.0 or greater
1 The uncertainty of the strength measurements is smallest when the soil conditions are uniform and high quality
strength test data provide a consistent, complete, and logical picture of the strength characteristics.
2 The uncertainty of the strength measurements is greatest when the soil conditions are complex and when availa
strength data do not provide a consistent, complete, or logical picture of the strength characteristics.
Duncan (1992), in a state-of-the-art paper on slope stability of soils, provided the following
discussion on the selection of an appropriate factor of safety:
"Criteria for acceptable values of safety factor should be established with two important
considerations in mind. These are, (1) what is the degree of uncertainty involved in
evaluating the conditions and shear strengths for analysis, and (2) what are the possible
consequences of failure? When the uncertainty and the consequences of failure are both
small, it is acceptable to use small factors of safety, on the order of 1.3 or even smaller in
some circumstances. When the uncertainties or the consequences of failure increase, larger
factors of safety are necessary. Large uncertainties coupled with large consequences of
failure represent an unacceptable condition, no matter what the calculated value of the
factor of safety. Typical minimum acceptable values of factor of safety are about 1.3 for end-
of-construction and multi-stage loading, 1.5 for normal long term loading conditions, and
1.0 to 1.2 for rapid drawdown, in cases where rapid drawdown represents an improbable or
infrequent loading condition."
While the guidance was developed by Duncan for soil slopes, the philosophy on FS selection is
directly applicable to the design of cover systems for waste containment applications and is
generally consistent with Table 6-3.
More recently, Koerner and Soong (1998) presented recommendations on FS selection that
incorporate a similar philosophy and that are specific to cover systems. The first step in the
approach suggested by Koerner and Soong (1998) is to qualitatively classify the project as
critical or non-critical and temporary or permanent. This qualitative classification is adapted
from Bonaparte and Berg (1987), who suggested its use for geosynthetic reinforcement in
highway applications. With this classification system, a critical application is one in which the
consequences of failure include a potential for personal injury, significant property damage, or
significant environmental release of contaminants. In contrast, a non-critical classification
would apply to a cover system of limited extent that could be readily repaired (e.g., a monolithic
soil cover) and for which the consequences of failure are minor. Table 6-4 presents the
qualitative classification system proposed by Koerner and Soong (1998). The classification
system is used to assign a ranking (low, moderate, or high) to the cover system so that the
appropriate FS value can be selected.
DRAFT - DO NOT CITE OR QUOTE
6-21
-------
Table 6-4. Qualitative classification for cover system applications (Koerner and Soong
(1998)).
Concern
Duration
Temporary Permanent
Noncritical
Critical
Low Moderate
Moderate High
Based on the cover system rankings in Table 6-4, Koerner and Soong (1998) recommended the
minimum static slope stability FS values for cover systems given in Table 6-5. Koerner and
Song (1998) suggested lower FS values for non-hazardous (principally MSW) landfills as
compared to hazardous waste landfills due to the differences in waste characteristics and the
larger magnitude of post-closure settlements for MSW landfills compared to hazardous waste
landfills (which results in an appreciable flattening of the MSW cover system slopes with time).
For Table 6-5, Koerner and Soong (1998) considered remediation waste piles to primarily consist
of low-hazard materials such as construction and demolition wastes and mine wastes.
Abandoned dumps on the other hand were considered to include CERCLA remediation sites and
other sites containing potentially-hazardous wastes or unknown wastes. Hence, abandoned
dumps were considered to pose a higher hazard than either non-hazardous waste landfills or
remediated waste piles.
Liu et al. (1997) have suggested that factors of safety for design of cover systems be selected by
a multi-step process that involves:
• estimating the mean value, standard deviation, coefficient of variation, and correlation
coefficient (between parameters) for each variable in the slope stability analysis;
• calculating failure probabilities and correlating these probabilities to the LE factor of
safety for the potential ranges in parameter values; and
• defining an acceptable probability of failure based on the cost and consequences of
failure.
Table 6-5. Minimum FS values for static slope stability of cover systems recommended
by Koerner and Soong (1998).
Type of Waste
Ranking
Low
Moderate
High
Remediated
waste piles
1.2
1.3
1.4
Non-hazardous
waste landfills
1.3
1.4
1.5
Abandoned
dumps
1.4
1.5
1.6
Hazardous waste
landfills
1.4
1.5
1.6
The approach described by Liu et el. (1997) provides a rational, probability-based approach to
designing safe cover system slopes. It is recognized, however, that not all engineers are
comfortable with probabilistic approaches and that the standard of practice is to use deterministic
DRAFT - DO NOT CITE OR QUOTE
6-22
-------
methods to establish factors of safety. At a minimum, however, Liu et al. (1997) provides a
useful framework for the design engineer to systematically consider the areas of uncertainty and
consequences of failure in the project.
A number of technical publications have addressed the issue of FS selection as it relates to the
use of peak versus large-displacement (or residual) internal or interface shear strength where the
large displacement strength of the material or interface is smaller than the peak shear strength
(Byrne, 1994; Stark and Poeppel, 1994; and Bonaparte et al., 1996).
The foregoing discussion should make it clear that there is no single value of FS applicable to all
situations. Selection of a FS value for a particular project is a key design decision that should be
the responsibility of the design engineer. Based on the foregoing discussion, the following
general guidance is given. This guidance applies specifically to cover systems, where the
geometry is well defined and the mass being analyzed consists entirely of manufactured or
constructed materials placed under controlled conditions. These minimum FS recommendations
may not be appropriate for other applications.
• A minimum acceptable factor of safety (FSm;n) for static stability analyses of 1.5 will
often be appropriate for permanent cover system applications where the design is based
on peak internal and interface shear strengths conservatively established using project-
specific interface direct shear tests, two-dimensional limit equilibrium slope stability
analyses, and appropriate consideration of the potential for internal hydraulic head build-
up during the representative design storm events. This FSm;n is applied to normal
operating conditions (e.g., no seismic loading or live loading).
• A smaller or larger FSmin may be considered based on an evaluation of: (i) consequences
of cover system failure; and (ii) uncertainty associated with each design parameter.
• If the cover system contains geosynthetic materials that exhibit strain-softening internal
or interface shear strengths, FSmin for large-displacement conditions should also be
checked. A FSmin of 1.2 is suggested where large-displacement shear strengths have been
conservatively established using project-specific interface direct shear tests conducted in
accordance with ASTM D 5321. For purposes of this evaluation, 50 to 75 mm of
displacement, coupled with the observation that the shear stress-displacement plot is
essentially flat at the end of the test, is considered to satisfy the large-displacement
condition. If true residual shear strengths are obtained using either a torsional-ring or
large-displacement shear apparatus, FSmin values as low as 1.15 may be considered.
• Cover system designs should be checked for low-probability extreme loading conditions.
These conditions need to be identified on a case-by-case basis, but may include extreme
storm events, live loads, or earthquakes. Design for earthquakes is addressed
subsequently. Values of FSmin for extreme loading conditions may, in general, be lower
than those associated with the representative design conditions as described above. FSmin
values for these conditions should be selected on a case-by-case basis.
If FSmin cannot be achieved for a given set of conditions, there are a variety of measures that can
be taken to increase its value. Examples of these measures are listed in Table 6-6.
DRAFT - DO NOT CITE OR QUOTE
6-23
-------
Table 6-6. Engineering measures to increase cover system slope stability factor of
safety.
Use cover system materials that have higher internal or interface shear strengths, as available
Provide for a flatter cover system slope by initially placing waste to a flatter slope (for new facilities)
or waste excavation (for existing facilities) (Figure 6-10)
Shorten the slope length through the use of benches or berms
Use perimeter retaining walls or buttresses to achieve a flatter cover system slope angle (Figure
6-11)
Improve cover system internal drainage if hydraulic head buildup is predicted to occur
Utilize geosynthetic reinforcement, but only within the limitations of this approach described in this
chapter
Original Slope-
Waste Excavation
Maximum slope
acceptable for cover
system installation
Figure 6-10. Waste Excavation Approach for Constructing Cover Systems Over Steep
Waste Slopes.
Maximum slope
acceptable for cover
system installation
Retaining
Wall
Wall
Height
Original
Slope
Figure 6-11. Buttress Approach for Constructing Cover Systems Over Steep Waste
Slopes Without the Need for Significant Waste Excavation.
DRAFT - DO NOT CITE OR QUOTE
5-24
-------
6.3 Seismic Slope Stability and Deformation
6.3.1 Overview
The cover system for a landfill or other waste containment unit or for a remediation site may be
subject to damage as a result of strong ground accelerations that can accompany an earthquake.
Impacts may involve either excessive seismic displacement of one or more of the cover system
components or complete instability of the cover system. For most situations, peak seismic
accelerations in a cover system will be larger than in the surrounding free field, due to
amplification of the ground movements by the underlying waste.
State and federal regulations have various requirements with respect to the evaluation of the
potential impact of seismically-induced ground motions on cover systems. EPA regulations for
hazardous waste landfills (40 CFR §264 and §265) are silent with respect to seismic design and
performance criteria. EPA seismic regulations for MSW landfills, contained in 40 CFR §258.14,
require that "all containment structures, including liners, leachate collection systems and surface
water control systems, are designed to resist the maximum horizontal acceleration in lithified
earth material for the site" if the landfill is located in a "seismic impact zone". EPA defines a
seismic impact zone as "an area with a ten percent or greater probability that the maximum
horizontal acceleration in lithified earth material, expressed as a percentage of the earth's
gravitational pull (g), will exceed 0. lOg in 250 years" EPA further elaborates that the
"maximum horizontal acceleration in lithified earth material means the maximum expected
horizontal acceleration depicted on a seismic hazard map, with a 90 percent or greater
probability that the acceleration will not be exceeded in 250 years, or the maximum expected
horizontal acceleration based on a site-specific seismic risk assessment." While this regulation
does not explicitly mention cover systems, EPA considers the cover system to be part of a
landfill "containment structure" and therefore covered by the regulation. However, the agency
recognizes that although difficult and potentially costly, cover systems can be repaired if
damaged. In contrast, landfill bottom liner systems generally cannot be repaired once covered
with waste. As a consequence, the agency believes that seismic performance criteria (e.g.,
acceptable FS or magnitude of permanent seismic deformation) applicable to cover systems may
not always need to be as stringent as those applied to landfill bottom liner systems.
The EPA guidance document entitled "RCRA Subtitle D (258) Seismic Design Guidance for
Municipal Solid Waste Landfill Facilities" (Richardson et al., 1995) presents available
information and analysis methods to evaluate the seismic performance of landfills. The
information contained in this guidance document is consistent with and includes new information
that has become available after publication of the EPA document listed above.
Evaluation of the seismic stability of a cover system involves four steps, each of which can be
performed using either conservative, simplified approaches, or more complex, detailed analyses.
These four steps, which are discussed in more detail below, are as follows:
1. conduct a seismic hazard evaluation to estimate peak horizontal bedrock accelerations for
a site and representative causative earthquake events to associate with that acceleration
(Section 6.3.2);
2. perform a seismic response analysis to evaluate peak horizontal accelerations at the
ground surface or in the waste mass cover system due to the causative earthquake events
DRAFT - DO NOT CITE OR QUOTE
6-25
-------
(Section 6.3.3);
3. select dynamic shear strength properties for cover system materials and interfaces to use
in seismic slope stability and/or deformation analyses (Section 6.3.4); and
4. perform seismic slope stability and deformation analyses (Section 6.3.5).
6.3.2 Seismic Hazard Evaluation
The objective of a seismic hazard evaluation is to characterize the design earthquake with respect
to the parameters required for engineering analysis (e.g., magnitude, style of faulting, site-to-
source distance, peak ground acceleration, and spectral accelerations). The peak horizontal
bedrock acceleration at a project site may be estimated using seismic hazard probability maps or
site-specific seismic hazard assessments. The most commonly used maps in the U.S. are those
developed by the U.S. Geological Survey (USGS) depicting peak and spectral horizontal bedrock
accelerations with 10, 5, and 2% probabilities of exceedance in 50 years (corresponding,
respectively, to a 90% probability of not being exceeded in 50, 100, and 250 years). These maps,
which can be downloaded from the USGS website (http://geohazards.cr.usgs.gov/eq/index.html),
are periodically updated to reflect recent developments in the field of seismology. Background
information on the development of these maps is provided by Frankel et al. (1996). Figure 6-12
presents the U.S. national map for peak horizontal acceleration in bedrock with a 90%
probability of not being exceeded in 250 years. A map for California and Nevada is presented in
Figure 6-13. These maps are included in this guidance document because the probability-
recurrence relationship for these maps corresponds to the EPA regulatory criterion for seismic
design of MSW landfills.
Seismic hazard maps like those of USGS discussed above usually present the estimated free-field
peak horizontal acceleration for a hypothetical bedrock outcrop on level ground at a particular
location. If bedrock is not present at or near the ground surface, the peak acceleration may need
to be modified to account for local site conditions. The presence of a waste mass will further
modify the earthquake ground motions, as discussed subsequently. The primary difficulty
associated with using seismic probability maps is that the maps by themselves do not provide
information on the magnitude, site-to-source distance, or duration of the earthquake associated
with the map acceleration values. For some types of seismic analyses, information on these
variables is necessary. Because they are probabilistically derived, the acceleration values
provided on such maps are composed of contributions of earthquakes of many different
magnitudes from several to many different seismic sources. Each source may be associated with
a different site-to-source distance and each magnitude-distance combination with a different
duration. The USGS website has recently made available information on the distribution of
earthquake magnitudes and site-to-source distances associated with the bedrock accelerations
obtained directly from the USGS seismic hazard maps. Using this feature, the peak bedrock
acceleration for a given site with a 2% probability of exceedance in 50 years (90% probability of
not being exceeded in 250 years) is deaggregated by earthquake magnitude and site-to-source
distance. Deaggregated spectral accelerations are also provided for spectral periods of 0.2, 0.3,
and 1 second. USGS currently provides deaggregated data for 64 central and eastern U.S. cities
and 56 western U.S. cities. As an example of the information available at the USGS website,
DRAFT - DO NOT CITE OR QUOTE
6-26
-------
Peak Acceleration (%g) with 2% Probability of Exceed an ce in 50 Years
site: NEHRP B-C boundary
-100
.70
U.S. Geological Survey
National Seismic Hazard Mapping Project
Figure 6-12. Seismic Hazard Probability Map for the U.S. for Peak Horizontal Acceleration in Bedrock with a 90% Probability
of not Being Exceeded in 250 Years (downloaded from http://geohazards.cr.usgs.gov/eq/).
DRAFT - DO NOT CITE OR QUOTE
6-27
-------
32
-724"
-120'
-118
-114
Figure 6-13. Seismic Hazard Probability Map for California and Nevada for Peak
Horizontal Acceleration in Bedrock with a 90% Probability of not Being
Exceeded in 250 Years (downloaded from http://geohazards.cr.usgs.gov/eq/).
DRAFT - DO NOT CITE OR QUOTE
6-28
-------
deaggregated bedrock acceleration and site-to-source distance data for Evansville, Indiana are
presented in Table 6-7.
In using deaggregated data, such as given in Table 6-7, the engineer should identify the
earthquake magnitude and distance combination that encompasses about two-thirds of the
seismic hazard. For example, if more than two-thirds of the seismic hazard for a given site is
from a small magnitude, near-source earthquake, then seismic analyses should be performed
using input variables (e.g., strong motion records) appropriate for this type of earthquake event.
In some cases, more than one combination of earthquake magnitude and source distance may
need to be considered. The values in Table 6-7 for Evansville illustrate such a case: a significant
portion (~ 40%) of the seismic hazard for Evansville is derived from earthquakes less than 25 km
from the site with magnitudes between 5 and 6 (though the magnitude of some of the local
events contributing to the seismic hazard may be as great as 7.5). However, over 20% of the
seismic hazard is from a distant earthquake more than 150 km from the site with a magnitude of
8.0. Therefore, for some projects in Evansville, the impact of both local and distant events may
warrant consideration.
Table 6-7. Deaggregated peak horizontal bedrock accelerations as a percentage of the
aggregated peak probabilistic acceleration of 0.328 g for Evansville, Indiana,
for a 2% probability of exceedance in 50 years (modified from USGS website).
Hypocentral
Distance
(km)
25
50
75
100
125
150
175
200
225
250
275
300
325
350
375
400
425
Earthquake Magnitude
5.0
13
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
888
507
093
011
002
000
000
000
000
000
000
000
000
000
000
000
000
5.5
13
3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
372
176
349
060
017
004
001
000
000
000
000
000
000
000
000
000
000
6.0
9.591
4.624
0.927
0.247
0.095
0.030
0.008
0.002
0.001
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
5
5
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
6.5
787
054
737
670
348
143
050
019
007
003
001
000
000
000
000
000
000
7.0
2.447
2.818
1.449
0.591
0.250
0.114
0.051
0.023
0.010
0.004
0.002
0.001
0.000
0.000
0.000
0.000
0.000
1
2
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
7.5
584
440
757
938
488
269
146
083
042
020
012
006
003
002
002
001
001
8.0
0.000
0.000
0.000
0.000
0.000
0.000
15.718
6.904
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
As a means to reduce uncertainty and increase accuracy, site-specific seismic hazard analyses
may be preferred to seismic hazard maps for assessing the seismic hazard to critical structures in
regions of high seismic activity. A site-specific seismic hazard analysis involves:
• identification of the seismic sources capable of strong ground motions at the project site;
• evaluation of the seismic potential for each capable source; and
• evaluation of the intensity of the design ground motions at the project site.
DRAFT - DO NOT CITE OR QUOTE
6-29
-------
Site-specific seismic hazard analyses may be performed using either a deterministic or
probabilistic approach. Detailed discussion of this topic is beyond the scope of this chapter. The
reader is referred to Reiter (1990), Krinitzsky et al. (1993), Richardson et al. (1995), and Kramer
(1996). An example of a site-specific seismic hazard analysis applied to a landfill site (including
cover system) in California is given in Kavazanjian et al. (1995a).
6.3.3 Seismic Response Analysis
6.3.3.1 Introduction
The seismic hazard assessment as discussed above provides an estimate of peak horizontal
accelerations in bedrock for a given site along with information on the causative earthquake
event(s). A response analysis is used to estimate the seismically-induced motions (e.g.,
acceleration, velocities, and/or displacements) at the ground surface or in the waste mass cover
system. Response analyses are needed because soil layers and waste modify the bedrock
motions, sometimes in a manner that can significantly increase damage potential.
6.3.3.2 Material Properties Selection
The first step in the seismic response analyses is to characterize the soil and waste material
properties needed to perform the analysis. For equivalent linear analyses of vertically-
propagating shear waves (the most common type of seismic response analysis performed for
geotechnical and waste management applications), these properties include total unit weight,
dynamic shear modulus, and damping ratio for each material through which the waves
propagate. Kramer (1996) provided an extensive review of the available technical literature on
the selection of soil and rock properties for response analyses. Guidance on selecting MSW
waste properties can be found in Sharma et al. (1990), Fassett et al. (1994), Richardson et al.
(1995), Kavazanjian et al. (1995b), Kavazanjian and Matasovic (1995), and Matasovic and
Kavazanjian (1998).
Shear modulus reduction factor and damping ratio curves for the Operating Industries Inc. (Oil)
site, a large inactive MSW and IW landfill in Monterey Park, California, were developed
independently by Idriss et al. (1995), Augello et al. (1998), and Matasovic and Kavazanjian
(1998). The curves proposed by these three sets of investigators are shown in Figure 6-14.
These curves represent the most reliable information currently available for use in estimating
strain-dependent dynamic shear modulus reduction factors, G/Gmax (dimensionless), and
damping ratios for MSW and other solid waste materials for use in equivalent linear response
analyses. The strain-dependent damping ratio is obtained directly from Figure 6-14. The
dynamic shear modulus, G (Pa), is obtained by multiplying the shear modulus reduction value
from Figure 6-14 by the maximum small-strain dynamic shear modulus, Gmax (Pa), which can be
calculated using the equation:
o
Y V
G = rt'waste s'waste (Eq. 6.16)
g
where: vS; waste = shear wave velocity of waste (m/s); yt, waste = total unit weight of waste (N/m3);
and g = acceleration of gravity (m/s2). The small-strain shear modulus is ideally obtained from
project-specific field testing. For landfills, this type of testing may be performed with the non-
DRAFT - DO NOT CITE OR QUOTE
6-30
-------
1.0
0.8
x 0.6
ro
E
CD
0.4
0.2
0.0
Idrissetal. (1995)
I
Augelloetal. (1998)
Matasovicand Kavazanjian (1998)
I
I
I
0.0001 0.001 0.01 0.1 1
Cyclic Shear Strain (%)
10
30
O
'ro
20
O)
c
'o.
03 10
Q
0
Matasovicand Kavazanjian (1998)
Idrissetal. (1995W
v
-Augelloetal. (1998)
I I I
0.0001 0.001 0.01 0.1
Cyclic Shear Strain (%)
1
10
Figure 6-14. Estimated Shear Modulus Reduction Factor and Damping Ratio Curves for
the ON Landfill, California (modified from Idriss et al.,1995; Augello et al,
1998; and Matasovic and Kavazanjian, 1998).
DRAFT - DO NOT CITE OR QUOTE
6-31
-------
Shear Wave Velocity (m/s)
100
200
300
400
500
600
0
15
30
Q_
0)
Q
45
60
75
I I
Mean plus 1 Std. Dev.
Mean minus 1 Std. Dev.
Mean
Profile recommended
for site response by
Kavazanjian et al.
(1995)
Recommended
range of values
for southern
California solid
waste landfills
Figure 6-15. Shear Wave Velocities for Southern California Solid Waste Landfills
(modified from Kavazanjian et al., 1996).
intrusive spectral analysis of surface waves (SASW) technique (Kavazanjian et al. 1994, 1996).
Intrusive downhole or cross-hole geophysical testing techniques may also be used. In the
absence of project-specific testing, the data for southern California landfills from Kavazanjian et
al. (1996), presented in Figure 6-15, can be used. It is noted that results obtained from a limited
DRAFT - DO NOT CITE OR QUOTE
6-32
-------
amount of SASW testing of MSW landfills in the eastern U.S. (unpublished) suggests that shear
wave velocities for waste in these facilities may be lower, on average, than shear wave velocities
for waste in relatively dry southern California landfills. In the absence of better information, the
lower portion of the recommended range of shear wave velocities shown in Figure 6-15 can be
used for MSW landfills located in the eastern U.S and other temperate to wet climates.
6.3.3.3 Simplified Response Analysis
Simplified approaches to seismic response analyses involve the empirical correlation of peak
horizontal waste mass or cover system acceleration, as applicable, to peak bedrock acceleration.
Correlations of this type were first used in geotechnical engineering to relate peak ground
accelerations at a site with a soil profile overlying bedrock to peak bedrock accelerations at the
same site (e.g., Seed and Idriss (1982) and Idriss (1990)). More recently, Bray et al. (1995),
Kavazanjian and Matasovic (1995), Singh and Sun (1995), Bray and Rathje (1998), and
Matasovic et al. (1998) have extended this type of relationship to solid waste landfills.
Matasovic et al. (1998) compared estimated horizontal bedrock accelerations to recorded peak
horizontal accelerations at the Oil site. Table 6-8, taken from Matasovic et al. (1998), presents
peak acceleration values (average of two horizontal components) recorded at the top deck of the
Oil landfill versus the estimated peak horizontal bedrock accelerations for the site. Based on
these results, Matasovic et al. (1998) concluded that peak horizontal bedrock accelerations from
both near-field and far-field earthquakes up to at least 0.15 g can be significantly amplified by
solid waste landfills. They suggested that, based on the Oil data, the curve developed by Harder
(1991) for the upper-bound amplification of seismic accelerations in earth dams, shown in Figure
6-16, provides a conservative upper bound for amplification of peak accelerations in solid waste
landfills. They also suggested that the relationship of Idriss (1990) for soft soil sites, shown in
Figure 6-16, provides a reasonable representation of average amplification potential of solid
waste landfills.
Table 6-8. Earthquake parameters, corresponding peak horizontal bedrock acceleration
estimates, and peak horizontal accelerations recorded on the top of the ON
Landfill, California (modified from Matasovic et al., 1998).
Earthquake Moment Style of Site-to-Source Estimated Peak Peak Acceleration
Magnitude Faulting Distance Bedrock Acceleration at Top Deck
(km) (g) (g)
Pasadena
(3 Dec 88)
Malibu
(19 Jan 89)
Joshua Tree
(23 Apr 92)
Landers
(28 Jun 92)
Big Bear
(28 Jun 92)
Mojave Desert
(11 Jul92)
Northridge
(17 Jan 94)
5.0
5.0
6.1
7.3
6.4
5.5
6.7
Strike-Slip
Thrust
Strike-Slip
Strike-Slip
Strike-Slip
Strike-Slip
Thrust
13
50
163
140
119
131
43
0.075
0.018
0.006
0.032
0.015
0.004
0.104
0.105
0.009
0.017
0.085
0.049
0.012
0.230
DRAFT - DO NOT CITE OR QUOTE
6-3
-------
0.6
0.5
0.4
(0
0)
O
< 0.3
^
o
Q.
.0
(0
0)
CL
0.2
0.1
0.0
/
/ I
//:'
/x
/
/
/
X
X
X
• /
/
X
larder (1
Earth Da
^'
X
\./
*y
391)
ms
-^
/ So
E
• N
a L
n P
A M
®|
A B
O IV
/
•iss (199
ft Soil Sit
arthquakc
/
3)
9S
^
orthridge
anders
asadena
alibu
Dshua Tree
g Bear
ojave Desert
0.0 0.1 0.2 0.3 0.4 0.5
Peak Bedrock Acceleration (g)
0.6
Figure 6-16. ON Data from Matasovic et al. (1998), Harder (1991) Curve for Upper-Bound
Amplification of Seismic Acceleration in Earth Dams, and Idriss (1990)
Curve for Soft Soil Sites.
o
O
0)
Q
(0
-------
Independent of Matasovic et al. (1998), Bray and Rathje (1998) used the non-linear one-
dimensional dynamic response analysis D-MOD (Matasovic, 1993; Matasovic and Vucetic,
1995) to perform parametric analyses of landfill response for a range of waste properties, waste
heights, site conditions, and bedrock ground motions. The results of their parametric evaluation
for cover systems are given in Figure 6-17. This figure presents a plot of peak horizontal
acceleration at the landfill top deck versus peak horizontal bedrock acceleration. Bray and
Rathje (1998) also compared their results to the Harder (1991) curve and the Seed et al. (1991)
curve for stiff soil sites. Inspection of Figure 6-17 shows that the Harder (1991) curve provides a
conservative upper bound to the calculated cover system accelerations. Bray and Rathje noted
that the large amount of scatter in their parametric analysis results is due in large part to the
sensitivity of the results to the input ground motion (i.e., variability among earthquakes).
Variability in the assumed foundation conditions and waste profile also influenced the results.
These findings are significant and engineers should consider this sensitivity when performing
and interpreting the results of seismic response analyses.
Until more data become available, the Harder (1991) curve is conservatively recommended as a
conservative upper-bound amplification of ground motions for simplified seismic site response
of solid waste cover systems. Knowing the peak horizontal bedrock acceleration (from a USGS
seismic hazard map, other map, or site-specific analysis), the Harder (1991) curve can be used to
estimate an upper-bound peak horizontal acceleration at the top deck of the landfill. Should the
Harder (1991) curve result in excessive cover system accelerations, detailed seismic response
analyses can be conducted to assess whether a lower value of peak acceleration can be used on a
project-specific basis. Site-specific seismic response analyses should also be used for any site
where the average shear wave velocity in the upper 30 m of the foundation is less than 120 m/s
(i.e., soft soil sites).
6.3.3.4 Analytical and Numerical Response Analyses
A one or two-dimensional seismic site response analyses may be performed for sites where
significant cover system accelerations are anticipated or it is necessary to obtain a better estimate
of seismically-induced motions in the cover system than can be obtained with the simplified
approach. These analyses are also recommended for sites with soft soil foundations and for
critical facilities or facilities with special features. Such projects include those in regions with
the potential for very large earthquakes, where waste thicknesses are relatively large, or where
cover system material or interface shear strengths are particularly low. The site response
analysis is performed considering both the foundation soils and waste mass.
The computer program SHAKE, originally developed by Seed and coworkers (Schnabel et al.,
1972) and updated by Idriss and Sun (1992), is perhaps the most commonly used computer
program for one-dimensional seismic site response analysis. The SHAKE model idealizes the
soil (and waste mass) profile as a system of homogeneous, visco-elastic sublayers of infinite
horizontal extent. The response of this system is calculated considering vertically propagating
shear waves. An equivalent linear analysis accounts for the strain-dependent non-linearity of soil
and waste stiffness and damping using an iterative procedure to obtain modulus and damping
values that are compatible with the equivalent uniform strain induced in each sublayer. At the
outset, a set of properties (shear modulus, damping ratio, and total unit weight) is assigned to
each sublayer of the soil or waste deposit. The analysis is conducted using these properties, and
DRAFT - DO NOT CITE OR QUOTE
6-35
-------
the shear strain induced in each sublayer is calculated. The shear modulus and damping ratio for
each sublayer are then modified based on the applicable relationship relating these two properties
and shear strain (see Figure 6-14).
Basic input to SHAKE includes the soil and waste profile, soil and waste properties, and selected
earthquake acceleration time history. Soil and waste properties include the shear wave velocity
(vs) or maximum (small-strain) dynamic shear modulus (Gmax) and total unit weight (yt) for each
soil layer plus shear modulus reduction and damping ratio curves for each soil and waste
material.
Computer programs are available for equivalent-linear and truly non-linear two and three-
dimensional seismic site response analyses. A discussion of these more sophisticated models is
provided by Kramer (1996). These models are only occasionally used in cover system design
practice. Application of these models to the evaluation of cover system earthquake response
often may result in lower-intensity seismically induced cover system motions than obtained
using the one-dimensional SHAKE analysis. Use of non-linear analysis methods is
recommended when the peak horizontal bedrock acceleration exceeds 0.4 g (Kavazanjian and
Matasovic,1995; Bray and Rathje, 1998). Examples of the use of these more sophisticated
models are presented by Idriss et al. (1995), Augello et al. (1998), and Matasovic and
Kavazanjian (1998), who used the two-dimensional finite element program QUAD4M (Hudson
et al., 1994) to evaluate the seismic response of the Oil landfill, and Kavazanjian and Matasovic
(1995), Bray and Rathje (1998),and Matasovic et al. (1998), who used the one-dimensional non-
linear program D-MOD (Matasovic, 1993) to evaluate landfill seismic response. These more
sophisticated models should only be applied by experienced geotechnical earthquake engineers,
as their application is quite complex.
To perform seismic site response analyses and/or to perform permanent seismic deformation
analyses, discussed subsequently, it is necessary to select earthquake acceleration-time histories
as an input parameter to the analyses. Acceleration-time histories can be developed either by
selecting a representative instrumental (accelerogram) record from the available catalog of
records obtained during previous earthquakes or by synthesizing an artificial accelerogram.
Acceleration-time histories should be selected for each seismic source having a potentially
controlling influence on a site. Both near-field and far-field earthquake events should be
considered. A higher magnitude far-field event with sufficient energy near the fundamental
period of a solid waste mass may be more damaging to an overlying cover system than a near-
field event characterized by a higher peak horizontal bedrock acceleration, higher frequency
motions, and a shorter duration. For analysis, each acceleration-time history is scaled to the peak
horizontal bedrock acceleration for the site. Selection of a time history from the available
catalog of time histories is, in general, a preferable approach as opposed to synthesizing a time
history. However, due to limitations in the catalog of available records, it is not always possible
to find a representative time history from the catalog, particularly for sites in the eastern U.S.
Richardson et al. (1995) and Kramer (1996) provide guidance on the selection of acceleration-
time histories for use in seismic analyses.
DRAFT - DO NOT CITE OR QUOTE
6-36
-------
6.3.4 Dynamic Shear Strength
The dynamic shear strengths of the components and interfaces of a cover system must be
estimated to perform seismic slope stability and/or deformation analyses. These estimates are
typically based on static or cyclic undrained shear strength tests. Shaking-table laboratory test
results and observed earthquake performance of cover system components and interfaces are also
used to develop information on cover system performance in earthquakes. Information on the
cyclic shear strength of soils used in cover systems can be obtained from the geotechnical
earthquake engineering literature (e.g., Kramer, 1996; Kavazanjian et al., 1997; and Lai et al.,
1998). Shear strengths of CCLs, dry GCLs, and unsaturated granular soils typically used in
cover systems are not significantly degraded by seismic loading, and cyclic shear strength is
assumed to equal static shear strength. For hydrated unreinforced GCLs, this may not be the
case, depending on the anticipated stress level and number of cycles of loading (Lai et al., 1998).
The limited available data on the cyclic shear strength of interfaces involving geosynthetics
(Kavazanjian et al., 1991; Yegian and Lahalf, 1992, Augello et al., 1995; Yegian et al., 1995; and
Chaney et al., 1997) suggest that cyclic shear strengths of geosynthetics can be approximated
using the results of static shear strength tests.
6.3.5 Seismic Stability and Deformation Analysis
6.3.5.1 Overview
The static LE slope stability analysis methods discussed previously in this document may be
adapted for use in the seismic stability evaluation of cover systems. This adaptation can be
achieved using a number of different approaches, of which the following three represent the
current state of practice: (i) the pseudo-static factor of safety method; (ii) the modified pseudo-
static factor of safety method; and (iii) the permanent seismic deformation method. These three
approaches are discussed below.
6.3.5.2 Psuedo-Static Factor of Safety Method
Due to its simplicity, the psuedo-static factor of safety method remains the most common
method of analysis used in practice for seismic design of cover systems. With this approach, the
factor of safety for the cover system is calculated using a LE analysis that incorporates a
specified seismic coefficient that is applied as a horizontal body force to the potential slide mass.
The factor of safety obtained for the calculation is compared to a minimum acceptable factor of
safety to determine the adequacy of the design. The seismic coefficient equals the fraction of the
weight of the potential failure mass that is applied as a horizontal force to the centroid of the
mass in a pseudo-static limit equilibrium stability analysis.
For the case of an infinite slope with no water flow, the pseudo-static factor of safety is given by:
(cosp-khsinp)tan4)i
kh cosp) ytt(sinp+ kh cosf
FS=
where: kh = pseudo-static seismic coefficient (dimensionless); and all other terms are as defined
previously.
DRAFT - DO NOT CITE OR QUOTE
6-37
-------
For the case of a slope of finite length, the pseudo-static factor of safety can be calculated, for the
case of no water flow, using the approximate solution for sliding of the two-part wedge shown in
Figure 6-4 (Matasovic et al., 2002):
FS = - • • a< • t
B ' Bytt 2hll-(B/A)tan
-------
of the peak horizontal acceleration of the slope (i.e., cover system). This result represents an
upper-bound value for the seismic deformations calculated by Hynes and Franklin (1984) using
almost 400 earthquake strong motion records. The value of kyg required to produce 0.3 m of
permanent seismic displacement drops to about 15% of the peak horizontal acceleration if the
mean plus one standard deviation curve is considered rather than the upper-bound curve. Other
values of kyg can be derived from Figure 6-18.
10'
E
-------
site, the mean plus one standard deviation ratios in Table 6-9 be used. Kavazanjian (1998)
recommended that seismic coefficients derived using Table 6-9 should be used with a factor of
safety of 1.0. It is cautioned that the use of peak shear strength parameters with this approach is
unconservative. Shear strength values should be selected considering the displacement value
from Table 6-9 associated with the chosen seismic coefficient. It is noted that this simplified
approach is not recommended for soft soil sites; soft soil sites should be evaluated using a site-
specific seismic response analysis and permanent seismic displacement analysis with
acceleration-time histories selected as previously described in this chapter.
Table 6-9. Ratio of yield acceleration, kyg, to peak acceleration of cover system as a
function of calculated permanent seismic displacement (based on Hynes and
Franklin (1984) curves shown in Figure 6-15). Note: a = standard deviation.
Calculated
Displacement (mm)
100
150
300
500
1,000
Mean
Ratio
0.23
0.17
0.08
0.05
0.03
Mean + 1a
Ratio
0.35
0.27
0.17
0.11
0.06
6.3.5.4 Permanent Seismic Deformation Method
With the permanent seismic deformation method, cumulative permanent seismic deformations
are calculated on the basis of that portion of the acceleration-time history of the cover system
that exceeds kyg. For the infinite slope case, ky is calculated using Eq. 6.17 and FS = 1.0. For
the case of a finite length slope with uniform soil thickness above the critical potential slip
surface, ky is calculated using Eq. 6.18 and FS = 1.0. For more complex cases, ky is calculated
using a LE slope stability computer program.
The actual calculation of permanent seismic displacement is usually performed using Newmark's
"sliding block on a plane" method of analysis (Newmark, 1965). In a Newmark analysis,
acceleration pulses (in the earthquake acceleration-time history) exceeding kyg are double-
integrated to calculate the accumulated "permanent" seismic displacement (Figure 6-19).
Theoretically, this calculated permanent displacement is a rigid body displacement that
accumulates everywhere along the critical potential slip surface. Typically, only the horizontal
component of the earthquake acceleration-time history is considered in the analysis. The
acceleration-time history of the cover system used in the analysis is obtained from a seismic
response analysis. With this approach, the response analysis is "decoupled" from the
computation of permanent displacement (i.e., seismic response is calculated assuming no slip
displacement between the cover system and landfill, and cover system displacement is calculated
using the results of the seismic response analysis (Bray and Rathje, 1998). The decoupled
approach is generally conservative for cover system displacement analyses.
Several commercially-available, PC-based computer programs exist to perform Newmark
analyses (Houston et al., 1987; Yan et al., 1996). These models assume a constant value of ky.
Recognizing that most geosynthetic materials and interfaces exhibit strain-softening shear
behavior, Matasovic et al. (1997) proposed a modification to the standard Newmark procedure
DRAFT - DO NOT CITE OR QUOTE
6-40
-------
Weight
Shear
Force
kh(t) Weight
Positive x Direction
Normal
Force
Time (t)
Figure 6-19. Basic Elements of Classical Newmark Sliding-Block Analysis with Constant
Yield Acceleration.
DRAFT - DO NOT CITE OR QUOTE
6-41
-------
Peak Parameters
Large Deformation or
Residual Parameters
Displacement, Q
(a)
en
>s
c.
.0
-------
o Newmark Analysis (Peak Parameters)
• Newmark Analysis (Residual Parameters)
A Newmark Analysis (Degradation Model)
0.0 0.2 0.4 0.6 0.8 1.0 1.2
Peak Horizontal Ground Acceleration (g)
Figure 6-21. Results of Newmark Seismic Deformation Analysis for Constant and
Degrading Yield Acceleration at a Normal Stress of 20.7 kPa (modified from
Matasovic et al., 1997). 81 and 82 are as Defined in Figure 6-20.
specifically for cover systems incorporating geosynthetic interfaces. The modified version
incorporates a linear ky degradation model to account for strain-softening materials and
interfaces (Figure 6-20). Matasovic et al. (1997) demonstrated the sensitivity of the calculated
permanent seismic deformation for a typical GT/CCL interface and three differing assumptions
regarding ky: (i) constant, based on peak interface shear strength parameters; (ii) constant, based
on residual (or large-displacement) interface shear strength parameters; and (iii) degrading, in
accordance with Figure 6-20. Figure 6-21 presents typical calculation results from Matasovic et
al. (1997) for the post-peak strain-softening exhibited by a GT/CCL interface. The sensitivity of
the calculation results to the ky assumption is evident.
6.3.5.5 Seismic Deformation Performance Criteria
In the current state-of-practice for design of cover systems, it is common to require permanent
seismic deformations calculated using a conservative, Newmark-type approach to be less than
150 to 300 mm (Seed and Bonaparte, 1992; Anderson and Kavazanjian, 1995). Smaller values
are sometimes considered if the potential slip surface underlies a non-ductile critical component,
such as a HDPE GM. Larger deformations are sometimes considered if the potential slip surface
is above all non-ductile critical components. Inherent in the selection of an allowable
displacement value is an understanding that the calculation methodology is conservative, and
actual earthquake-induced deformations would be less than calculated. In this regard, some
DRAFT - DO NOT CITE OR QUOTE
6-43
-------
engineers prefer to view the calculated seismic displacement as a performance index as opposed
to a true prediction of actual deformations.
In applying seismic performance criteria to cover systems, several factors can be considered that
do not typically apply to liner systems. First, the condition of a cover system can be readily
observed after an earthquake through a post-earthquake inspection program. Second, the
potential adverse impacts associated with excessive deformation of a cover system will involve
tearing of geosynthetics, cracking of soils, disruption of gas management systems, and disruption
of surface-water management systems. The risk of personal injury or environmental impact
resulting from these types of problems will typically be small. The damage to cover systems
caused by seismic displacement is typically repairable, although some at considerable cost and
effort. For these reasons, it may be acceptable in some cases to consider calculated permanent
displacements that are near the upper limit of the current state-of-practice for cover system
applications.
Kavazanjian (1998) proposed two criteria for the seismic design of cover systems: (i) design
without damage; and (ii) design accepting some limited damage to the cover system, but without
"harmful discharge." For the "no damage" criterion, Kavazanjian suggested that a calculated
permanent seismic displacement of up to 300 mm is acceptable for simplified analyses which use
upper bound displacement curves from generic Newmark displacement charts (e.g., Hynes and
Franklin, 1984), residual shear strengths, and/or simplified seismic response analyses.
Kavazanjian further suggested that a calculated permanent seismic displacement of up to 150
mm represents an acceptable "no damage" criterion in cases where more sophisticated analyses
are used to calculate the permanent seismic displacement using project-specific seismic response
and formal Newmark displacement analyses.
For the case of "no harmful discharge", Kavazanjian (1998) suggested that a permanent
deformation criterion of up to 1 m may be acceptable. With respect to this criterion, Kavazanjian
(1998) states:
"When designing a cover system to withstand [an earthquake] without discharge, provisions
are needed to mitigate potential hazards associated with discharge ofleachate and/or gas
from disrupted conveyance systems (e.g., use of automatic shut-off valves, secondary
containment, and/or articulated seismic joints) and facilitate post-earthquake repair of
damage."
"Multiple penetrations through geomembrane cover elements for gas and leachate collection
or other purposes may limit allowable displacement to less than 1 m on an economic basis
due to the cost of repair. However, if the anticipated displacement is above the
geomembrane, there are not penetrations through the geomembrane on slopes, and benches
provide sufficient capacity to retain cover soil that sloughs from above, the allowable seismic
displacement of a geosynthetic landfill cover system may be unlimited, provided the owner is
prepared to repair and/or replace the protective soil cover and drainage layer (if any) on top
of the geomembrane after a severe earthquake."
The choice of a "no damage" or "no harmful discharge" design criterion will need to be made on
a case-by-case basis by the design engineer, facility owner, and regulatory agency. Obvious
DRAFT - DO NOT CITE OR QUOTE
6-44
-------
factors that should be considered in choosing a criterion are: (i) potential impacts of large
displacements; (ii) type of waste being covered; (iii) cost to repair cover system damage; and (iv)
level of assurance that personnel and funds will be available for post-earthquake inspections and
repairs after the earthquake occurs.
6.4 Settlement
6.4.1 Mechanisms of Settlement
Cover systems may be subject to settlements resulting from a variety of mechanisms. For
purposes of evaluating cover system performance, settlements can be considered to have one of
three sources (see Figure 6-22): (i) settlement of foundation soil; (ii) settlement due to overall
waste mass compressibility; and (iii) settlement due to localized mechanisms in the waste.
Angular distortion or differential settlement may: (i) induce unacceptable tensile stress and strain
in one or more cover system components, which can lead to component tearing or cracking; or
(ii) cause cover system slopes to change or reverse grade which, in turn, can affect the
performance of the cover system drainage layer and/or gas collection layer.
Overall Cover System—
Settlement
Localized Cover System ,
Settlement
Original __
Slope \,~***'
_ J
Highly-Compressible
Waste Material
Figure 6-22. Sources of Cover System Settlement (modified from Othman et al., 1995).
6.4.2 Settlement of Foundation Soils
Impacts of foundation settlement on the performance of a cover system are usually not
significant. Occasionally, situations arise where foundation settlements are of sufficient
magnitude to affect the cover system design. For example, if the waste mass is underlain by a
thick layer of soft clay, consolidation settlements can be large. Both primary settlement and
long-term secondary settlement should be considered. Calculations are performed using
equations from conventional geotechnical engineering practice (e.g., Holtz and Kovacs, 1981;
DRAFT - DO NOT CITE OR QUOTE
6-45
-------
Lambe and Whitman, 1969) and a timeframe at least equal to the active life and post-closure care
period of the facility.
6.4.3 Overall Waste Compression
Overall waste mass compressibility results in area-wide waste mass settlement. The potential for
waste settlement is highly dependent on the type of waste. Relative to most other wastes, MSW
is very compressible, due to both its initial compressibility when placed and the additional
compressibility induced by the biodegradation of the organic component of the MSW. This
latter component creates a significant time dependency to waste settlement. Other types of
facilities that can undergo large settlement include impoundments containing high water content
industrial sludges (typically inorganic). Materials such as mine waste, ash and slag, construction
and demolition waste, and soil waste have relatively lower settlement potential. The following
discussion of overall waste settlement focuses primarily on the settlement potential of MSW
waste and other highly compressible waste material. The evaluation of ash, soil waste, or other
low-compressibility inorganic waste is typically performed using equations for conventional
geotechnical engineering practice (e.g., Holtz and Kovacs, 1981, Lambe and Whitman, 1969).
MSW waste compression results from complex factors including (Sowers, 1973; Edil et al.,
1990; Sharma and Lewis, 1994):
• mechanical compression due to self-weight and surface loads;
• raveling (i.e., movement of fines into larger voids);
• physiochemical changes, including corrosion, oxidation, and combustion; and
• biochemical decomposition under aerobic and anaerobic processes.
The magnitude and rate of MSW settlement are controlled by many factors, among which are the
waste fill height, organic content, age, moisture content, degree of compaction, and temperature.
Figure 6-23 presents data from Edgers et al. (1992), Konig et al. (1996), and Spikula (1996),
which shows that MSW landfills can settle from about 5 to 20% (and even up to 30%) of the
initial landfill thickness (measured from the time the landfill first reached final grade).
A number of methods have been proposed for evaluating the short-term and long-term
compression of waste. Three settlement models that have been adopted from geotechnical
engineering and applied to waste are: (i) one-dimensional compression model; (ii) power creep
model; and (iii) Gibson and Lo model (Gibson and Lo, 1961). A discussion of the latter two
models is presented in Edil et al. (1990), and they are not discussed further herein. Presently,
there is little experience in applying these last two models, and their applicability to the
prediction of long-term settlements is not well demonstrated.
Conventional one-dimensional compression models have been widely used to estimate waste
settlements (Sowers, 1973; Yen and Scanlon, 1975; Rao et al., 1977; Burlingame, 1985; Landva
and Clark, 1990; Morris and Woods, 1990; Fassett et al., 1994; Stulgis et al., 1995). However, it
is often assumed that primary self-weight settlement is complete prior to installation of the cover
system. Thus, it is assumed that calculated primary settlements do not directly influence cover
system performance.
DRAFT - DO NOT CITE OR QUOTE
6-46
-------
CO
CO
-------
some MSW materials, Cae is more or less constant during the period for which data exist, while
for other facilities, a variable Cae could be used to better fit the data.
The reader should be aware that the choice of a value of Cae cannot be made without
consideration of the normalization term ti. For a given Cae, the calculated value of AHS will be
significantly affected by the choice ofti. Ideally, Cae and ti should be selected by empirically
fitting Eq. 6.19 or Eq. 6.20 to field settlement data. In the absence of this type of correlation, it is
suggested that ti be taken as the time period between when waste reaches final grade and when
the cover system is installed over the waste.
Since Cae and ti are empirically derived, AHS is assumed to be independent of applied effective
stress, and the primary purpose of calculating AHS herein is to assess potential impacts to the
performance of the cover system, it is not necessary to subdivide the waste mass into a series of
horizontal layers for purposes of calculating AHS. With this approach, calculations are typically
performed for increasing time intervals after closure to obtain a relationship between cover
system settlement and elapsed time since closure.
Values of Cae for MSW reported in the technical literature have generally been in the range of
0.01 to 0.1 (Sowers, 1973; NAVFAC, 1983; Burlingame, 1985; Landva and Clark, 1990; Fassett
et al., 1994; Stulgis et al., 1995). Given the empirical nature of Cae and ti, it is interesting to
compare calculated values of (AHs/Hi) obtained using Eq. 6.19 to the range of observed time-
dependent post-closure settlements for MSW (Figure 6-20). For the comparison, the waste mass
is considered as a single unit with an average ti value of 100 days (approximately 3 months).
Table 6-10 presents calculated values of AHs/Hi (in percent) for values of Cae ranging from 0.01
to 0.1 and post-closure times of 100, 1,000 and 10,000 days (to which 100 days are added to
obtain t2 values).
Table 6-10. Results of parametric study of calculated post-closure secondary
settlements (AHS) as a percentage of initial landfill height (Hi).
:i::
0.01
0.03
0.06
0.10
100
0.30
0.90
1.8
3.0
(t2 - t|) (days after closure)
1,000 10,000
1.0 2.0
3.1 6.0
6.2 12.0
10.4 20.0
Based on the calculated values in Table 6-10, Cae values less than about 0.03, coupled with ti
values of 100 days, are too small to model MSW time-dependent settlements. Careful review of
the source references used to develop Figure 6-23 suggests that Ca£ values in the range of 0.04 to
0.08, coupled with ti values of about 100 days, provide a reasonable modeling of the settlement
trends for modern MSW landfills that are typically filled fairly quickly and compacted using
heavy trash compactors. Larger values of Cae, in the range of 0.08 to 0.12, coupled with ti = 100
days, are needed to model the settlement trends in some of the older landfills in the source
database. These landfills may have been filled with more variable waste placed under conditions
less controlled than for modern landfills. Larger values of Cae would also be expected for
DRAFT - DO NOT CITE OR QUOTE
6-48
-------
landfills undergoing leachate recirculation or otherwise managed to increase biological activity
and methane production in the waste mass.
If ti is assumed to be 30 days rather than 100 days, calculated AHs/Hi values at 10,000 days
would be about 25% larger than given in Table 6-10. Thus, if ti = 30 days is assumed, Cae values
should be reduced about 25% from the recommended ranges given above. This calculation
exercise clearly points out the sensitivity of calculated AHs/Hi values to the magnitude ofti.
6.4.4 Differential Settlement Due to Localized Mechanisms
Localized settlements, in the form of depressions, can develop within the first several years after
cover system installation over MSW. These types of localized occurrences appear to be more
common in older waste fills where a number of factors may contribute to the problem, including:
(i) little initial waste compaction; (ii) variable waste characteristics; (iii) placement of sludges in
the waste fill; and/or (iv) poor surface-water control leading to ponding of water on the waste.
Localized differential settlement can lead to excessive stress or strain in cover system
components (Gilbert and Murphy, 1987). Localized differential settlement of waste is generally
attributed to one or more of several mechanisms, namely: (i) deterioration and collapse of objects
(e.g., drums) in the waste; (ii) settlement associated with a highly-compressible zone of waste;
and (iii) migration or raveling of waste particles into underlying voids.
Typically, analyses to evaluate impacts of localized differential settlement on the cover system
are not performed as part of cover system design. However, in a few situations it may be
necessary to evaluate potential effects of localized areas of high waste compressibility on cover
system performance (e.g., cover systems for old dumps where the composition of waste is
unknown or there is reason to believe that significant local waste heterogeneity may exist (due to
any of the factors described above)). Several analysis methods are available for use in evaluating
the potential effects of localized settlements on cover system performance. None of the methods
have been field calibrated to any significant degree and selection of input parameters to the
analyses is based primarily on engineering judgment. The analysis methods include:
• the application of mine subsidence models to the prediction of waste differential
settlements (Murphy and Gilbert, 1985);
• an approach based on the uncoupled combined use of the tensioned membrane and soil
arching theories for analyzing the stresses and strains in geosynthetics (such as
geosynthetic layers within a cover system) that lose foundation support after construction
due to development of a foundation void or depression (Giroud et al., 1990);
Poorooshasb (1991) used a somewhat different analytical approach to address a similar
problem;
• a boundary element formulation to model deformations around a collapsing void within
an existing waste mass (Jang and Montero, 1993);
• two-dimensional finite element analyses to evaluate the response of a waste mass
containing compressible zones (Carey et al., 1993); and
• the displacement method of Sagaseta (1987) to evaluate the response of a cover system
over a waste mass containing localized compressible zones (Othman et al., 1995)
DRAFT - DO NOT CITE OR QUOTE
6-49
-------
In the situation where differential settlement is likely to occur and the localized depressions
cannot be eliminated, the choices are (i) continuously grading and maintaining the site; (ii)
installing a thick buffer soil or waste prior to cover system construction; or (iii) installing
geosynthetic reinforcement beneath the cover system. One or more of the analysis methods
described above can be used to design soil buffer or geosynthetic (geogrid or high strength GT)
support systems. The critical design parameters in any such analysis are the locations and
dimensions of the anticipated localized depression or void. Since it is generally not possible to
predict where such a feature will occur, any buffer soil or reinforcement layer, if used, will
typically need to be installed over the entire waste mass.
6.4.5 Impacts of Settlement on Cover Systems
In design, settlement profiles accounting for the various settlement mechanisms are developed to
evaluate potential impacts to the cover system. The evaluation usually considers: (i) post-
settlement cover system grades; (ii) potential for depressions and ponding in the cover system;
and (iii) stresses and strains in cover system components. Post-settlement grades should be
adequate to shed runoff, prevent ponding, and prevent excessive stress or strain in cover system
components, particularly the CCL, GCL, and GM hydraulic barriers.
Tensile strains causing cracking in compacted clays have been evaluated by Leonards and Narain
(1963); Ajaz and Parry (1975a,b, 1976); Gaind and Char (1983); Chandhari and Char (1985);
Jessberger and Stone (1991); and Lozano and Aughenbaugh (1995). Based on these studies,
compacted clays tested under unconfined or low confinement conditions exhibit relatively brittle
behavior and reach failure at axial extensional strains of 0.02 to 4%, with most compacted clays
exhibiting failure at extensional strains of 0.5% or less. The studies also showed that the
magnitude of tensile strain causing cracking increases with increasing percentage of fines and
water content, and with increasing confining stress.
LaGatta et al. (1997) evaluated the impact of differential settlement on the hydraulic conductivity
of GCLs overlain by a 0.6-m thick layer of pea gravel. The GCLs were tested either dry or
hydrated and either intact or with a 230 mm overlap. The overlapped GCLs were tested across
the overlap. The angular distortions (see Figure 2-13) of the upper surface of the GCLs were
monitored and used to calculate tensile strain. The results of their evaluation indicate that intact
and overlapped samples of needlepunched GCLs can withstand angular distortions of 0.35 to 0.6,
equivalent to tensile strains of 5 to 16%, while maintaining a saturated hydraulic conductivity of
1 x 10"9 m/s or less. Stitch-bonded GCL samples were found to achieve the same hydraulic
conductivity criterion up to an angular distortion of 0.35, equivalent to a tensile strain of 5%.
For GT-encased, non-reinforced GCL samples, the maximum allowable angular distortion was
much less, only 0.1, which is equivalent to a tensile strain of about 1%. This type of GCL, which
is no longer available, had an open weave GT on its lower surface. The GT provided essentially
no support to the GCL and allowed bentonite to migrate downslope within the depressed area
and experience significant swelling. At the end of the test, the thickness of hydrated bentonite
was approximately 5 mm on the sides of the depression and 50 mm on the floor of the
depression. GCLs samples consisting of bentonite adhered to a GM maintained an equivalent
hydraulic conductivity of 1 x 10"9 m/s or less when subjected to angular distortions of up to 0.8,
producing a maximum tensile strain of almost 30%. Migration of bentonite was not observed
DRAFT - DO NOT CITE OR QUOTE
6-50
-------
because the GM component of the GCL blocked most of the flow. Within the test area, the GCL
was only hydrated along the part of the overlap.
The tensile behavior of GMs varies depending on the polymer type, stress-strain characteristics,
susceptibility to stress cracking, temperature, and other factors. The present state-of-practice for
the design of strain-softening GM barriers (e.g., polyethylene GMs) is to limit the allowable GM
tensile stress (or strain) to the short-term yield value divided by a factor of safety. The allowable
tensile stress (or strain) for GMs exhibiting strain-hardening behavior (e.g., PVC GMs) is based
on the short-term failure (break) value divided by a factor of safety. It should also be
remembered that GMs are designed to be barriers, not tensile inclusions (as is geosynthetic
reinforcement, for example). The long-term stress-strain, creep, and brittle fracture behavior of
these materials under stress is not well understood. To the extent possible, applications should
be designed to minimize tensile stresses and elongations in GMs.
The authors recommend that when it is necessary to specify allowable geosynthetic tensile
stresses and strains, the yield stress and strain of the GM material be determined in a wide-width
tension test (for plane deformation) or axisymmetric tension test (for spherical deformation) and
that the factor of safety used to calculate the allowable values be at least five. The factor of
safely should be applied to the yield values for strain-softening GMs and to the failure (break)
values for strain-hardening materials. This recommendation should be conservative for virtually
all types of commercially-available GMs used in cover system applications. If a higher value of
allowable tensile stress or strain is desired, the design engineer must demonstrate that the product
to be specified can sustain the allowable values without long-term creep, brittle rupture, or other
type of long-term problem. This demonstration must also apply to GM seams.
6.5 Steep Slopes
6.5.1 Introduction
Occasionally, in the closure of old, existing landfills, it is necessary to address the issue of steep
existing waste slopes. One option is to cut the slope back to a shallower grade by excavation and
then relocate the excavated waste either on-site or off-site at another landfill (Figure 6-10). The
advantage of this approach is that it increases the stability of the waste mass and results in a final
slope inclination within the "conventional" range for cover systems. Disadvantages associated
with waste excavation and relocation include construction-related instability, health and safety
concerns associated with exposing the waste, leachate generation, nuisance (e.g., odor) issues,
waste characterization necessary for on-site or off-site disposal of the excavated waste, and cost.
Several alternative approaches exist for constructing cover systems over steep waste slopes
without need for waste excavation, or at least with very limited waste excavation. These
alternatives include the use of: (i) a waste buttress coupled with a conventional slope cover
system (Figure 6-11); or (ii) a steep slope cover system. Both of these alternatives are illustrated
below, primarily in the form of case studies illustrating their use.
6.5.2 Waste Buttress
Two examples of the use of a waste buttress in the closure of old, existing landfills are presented
below in order to illustrate the application of this technology.
DRAFT - DO NOT CITE OR QUOTE
6-51
-------
Cargill and Olen (1997, 1998) describe the closure of an inactive hazardous waste landfill
located on Long Island, New York. The landfill covers approximately 19 ha and has waste
slopes extending up to 42 m above the surrounding ground surface. The steepness of the existing
waste slopes, with inclinations up to 1H:2V and an average inclination steeper than 2H: IV,
prevented the use of a conventionally-designed cover system. Regrading of the landfill to
achieve slopes on which a conventional cover system could be constructed was not feasible due
to limits on the final landfill height and lack of alternate landfilling locations for the excavated
waste. For this project, the cover system design for the steeper slope sections incorporated a
Reinforced Soil
Slope Face Wrap
Vegetated
Top Soil
0.46m
Varies
Secondary Soil
Reinforcement
Primary Soil
Reinforcement
GM
- Gas Transmission GC
Figure 6-24. Detail of Reinforced Soil Slope Cover System Used on Steeper Slope
Sections of a Hazardous Waste Landfill Cover System (modified from Cargill
and Olen, 1997, 1998).
shingled GM within a geogrid-reinforced waste buttress (Figure 6-24). Approximately 4,300
linear m of slope buttress was constructed at heights up to 6.1m. A cross section of this
buttressed cover system is shown in Figure 6-25, and photographs of the system during
construction and after completion are shown in Figures 6-26 and 6-27.
Slope stability is a major factor in the design of a cover system such as that described above.
Three broad types of stability conditions must be considered for this type of closure. The first
condition involves internal and interface stability of the components of the conventional portion
of the cover system. The stability of these conventional components are evaluated using the
procedures described in Sections 6.2 and 6.3 of this document.
DRAFT - DO NOT CITE OR QUOTE
6-52
-------
60
54
Conventional
Cover System
48
L42
>
•36
_
LU
Waste Fill
Reinforced
Soil Slope
Cover System
EXISTING GROUND
Additional Fill
Waste Excavation
U)
>i O
£=
en a)
EXISTING WASTE
30m
60 m
90 m
150m
Figure 6-25. Reinforced Soil Slope Cover System on Steeper Slope Sections and
Conventional Cover System on Shallower Slope Sections of a Hazardous
Waste Landfill (modified from Cargill and Olen, 1997, 1998).
Figure 6-26. Construction of Reinforced Soil Slope Cover System on Steeper Slope
Sections of a Hazardous Waste Landfill.
DRAFT - DO NOT CITE OR QUOTE
6-53
-------
Figure 6-27. Constructed Reinforced Soil Slope Cover System for a Hazardous Waste
Landfill.
The second condition involves the internal stability of the waste buttress. Many different types
of earth retaining structures, including crib walls, mechanically stabilized earth (MSB) walls, and
reinforced soil slopes, can be used in this application. If the structure is to be founded on firm
ground, it can be fairly rigid, such as a precast concrete bin wall. However, if the structure is to
be founded on waste, it must be flexible and able to undergo significant settlement and distortion
while maintaining functionality. Geosynthetic-reinforced MSB walls and slopes with flexible
facing elements meet these criteria. Cargill and Olen (1997, 1998) utilized geogrid-reinforced
soil to form the buttress component of the cover system and a flexible facing (Figure 6-28).
Procedures for design of earth retaining structures and evaluation of the internal stability of these
structures can be found in a series of documents published by the U.S. Department of
Transportation Federal Highway Administration (FHWA) (i.e., Holtz et al., 1995; Elias and
Christopher, 1996; Sabatini et al.,1997) and in geosynthetic textbooks (Koerner, 1998).
The third stability condition that must be considered is the global stability of the buttress, waste
mass, and landfill foundation. Global stability is typically evaluated using classical two-
dimensional, LE slope stability analysis methods (e.g., Bishop, 1955; Spencer, 1967;
Morgenstern and Price, 1965), as coded in the previously-mentioned commercially-available,
PC-based computer programs (see Section 6.2.2.3). A critical aspect in the evaluation of global
stability is the establishment of shear strength and unit weight parameters for soil and waste
materials, and liquid heads (e.g., leachate heads in the waste and/or groundwater heads in the
DRAFT - DO NOT CITE OR QUOTE
6-54
-------
Varies
5.5ft
Seeded
Erosion
Control
Blanket
Topsoil
-Secondary Reinforcement
-Structural Fill
1.5ft
^—Primary Reinforcement
Figure 6-28. Flexible Facing Used with Reinforced Soil Slope Cover System for a
Hazardous Waste Landfill (modified from Cargill and Olen, 1997; 1998).
foundation). Shear strengths for soil materials can be established using project-specific
geotechnical site investigations and laboratory testing programs. For waste, shear strength and
unit weight parameters can be established from information in the technical literature (e.g.,
Landva and Clark, 1990; Fassett et al., 1994; and Kavazanjian et al., 1995b) or through the use of
project-specific field and laboratory test programs. For liner system materials and interfaces,
shear strength parameters can be established using a laboratory testing program that includes
consideration of the relevant testing procedures discussed for cover systems in Section 6.2.4. If
the potential slip surface passes through a strain-softening soft soil foundation or liner system
interface, the shear strength values selected for the waste, foundation, and/or liner system must
be based on strain compatibility between the various materials along the potential slip surface.
For example, the shear strain necessary to develop the peak shear strength of MSW may
correspond to post-peak (e.g., residual) shear strength of a soft soil foundation material.
For another project, Graves et al. (1998) discussed the use of a pre-cast concrete crib wall 915m
long and up to 9 m high as part of an upgraded closure activity and flood protection for a 20-ha
unlined sanitary landfill in Cuyahoga County, Ohio. In the years after landfill closure, an
adjacent creek caused erosion and undermining of the landfill, creating localized near-vertical
waste slopes (and overall waste slopes of about 2H:1V) and causing concerns about overall
stability of the landfill. Flattening of the slope to allow installation of a conventional final cover
system and achieve adequate slope stability factors of safety would have required relocation of
approximately 700,000 m3 of waste. Through use of a crib wall buttress at the toe of the landfill,
the required waste excavation volume was reduced to 280,000 m3. Design details and
construction photographs for this project are illustrated in Figures 6-29 through 6-32.
The case study presented above utilized a pre-cast concrete crib wall as a gravity buttress to
support waste slopes. Other types of wall systems could also be used for this application.
Reference should be made to Sabatini et al. (1997) for an inventory of available wall types and
typical wall unit costs. An emerging technology that holds promise for future use in landfill
stabilization projects involves the use of geofiber reinforcement. Geofibers consist of relatively
DRAFT - DO NOT CITE OR QUOTE
6-55
-------
37m
(typ)
-Gas Extraction
Well
Original Configuration
of West Slope Toe
(undermined by Creek)
3-6 m
(typ)
v :
Gas Collection
Header Pipe
WASTE
Pre-construction
West Slope
2 (typ) Ns
'—Shale Foundation
- Pre-construction
West Slope
Desired2.5H:1V
Slope Configuration
WASTE
Access Road
With Riprap
Protection
Shale Foundation
Figure 6-29. Waste Buttress Reduced Waste Excavation Volumes Required for an
Upgraded Closure Activity at a Sanitary Landfill (modified from Graves et
al., 1998).
DRAFT - DO NOT CITE OR QUOTE
6-56
-------
Security Fence
Crib Wall
Vegetated Topsoil Layer
CCL
Pre-Construction West Slope
WASTE
Access Road
Riprap
Protection
PVCGM
GCL
Mill Creek
GT Filter
Wall Drain And Leachate
Collection System
Soil-be ntonite
Cut-off
Figure 6-30. Pre-cast Concrete Crib Wall Used as Waste Buttress for a Sanitary Landfill
(modified from Graves et al., 1998).
Figure 6-31. Construction of the Pre-cast Concrete Crib Wall Waste Buttress for a
Sanitary Landfill.
DRAFT - DO NOT CITE OR QUOTE
6-57
-------
Figure 6-32. Aerial View of Constructed Cover System with a Pre-cast Concrete Crib Wall
Waste Buttress at a Sanitary Landfill.
small (e.g., 25 to 50 mm in typical length) polymeric inclusions, distributed throughout the
reinforced soil mass. There are a variety of techniques for mixing the fibers into a soil fill,
including pneumatic application and mechanical mixing. An example of the use of geofibers for
a slope stabilization project is given in Gregory and Chill (1998).
6.5.3 Steep Cover System Slopes
Cover system slopes somewhat steeper than those conventionally used can be achieved through
the careful selection of cover system components. Materials that can be used to increase the
inclination of cover system slopes include:
• textured GMs or GMs manufactured from polymers that generate higher interface shear
strengths compared to smooth GMs manufactured from HDPE;
• geosynthetic reinforcement installed parallel to the landfill slope in the internal drainage
layer or protection layer and anchored at the crest of the slope;
• geosynthetic drainage layers;
• geofiber reinforcement of cover system soil layers;
• geocell (e.g., Figure 6-33) and geosynthetic erosion control materials (described in
Section 2.2.5.3); and
DRAFT - DO NOT CITE OR QUOTE
6-58
-------
• topsoil surface/protection soil layer with adequate cohesion to resist erosion, yet with
adequate characteristics to support vegetative cover.
(a) Expanded
230 mn
2.4m
230 mn
(b) Collapsed
3.4m
ftt *I ,
Figure 6-33. Geocells can be Used to Reinforce the Topsoil Surface/Protection Layer on
Steep Cover System Slopes.
DRAFT - DO NOT CITE OR QUOTE
6-59
-------
Using these materials, cover system slopes as steep as 2.5H:1V, and possibly slightly steeper can
be constructed and maintained at a factor of safety at or near the target range discussed
previously in this chapter. Even steeper slopes (e.g., 2H: IV, as demonstrated in the GCL test
plot program described in Chapter 7 of this document) can be constructed, but factors of safety
are likely to be lower than the target range given in herein. Moreover, long-term surface erosion
problems should be anticipated when steep slopes are used. With steeper slopes, several other
design aspects take on even greater importance than they might otherwise. For example, greater
attention must be given to the selection of internal and interface shear strengths due to the greater
potential for slope instability; thus, project-specific shear strength testing is essential. Also,
seepage in a slope containing geosynthetic reinforcement can greatly reduce the effectiveness of
the reinforcement. The stress-elongation characteristics of the various cover system components
also become more important as slopes become steeper. Thus, the consideration of deformation
compatibility of the cover system components is essential. It is possible, for example, that the
elongation required to induce the design tensile force in a geosynthetic reinforcement layer is
larger than the shear deformation needed to cause a GCL to exhibit large displacement rather
than peak internal shear strengths characteristics. For this case, the design should be based on
the large-displacement GCL shear strength and not the peak shear strength. Deformation
compatibility can be evaluated as previously discussed in Section 6.2.3.
6.6 Soft Waste Materials
Another type of design issue sometimes encountered is the in-situ closure of impoundments or
the capping of remediation source areas containing soft waste materials. These soft materials
include high moisture content sludges, saturated process wastes, and saturated sediments or solid
wastes. The common characteristic of these materials is that they have very low bearing
capacity, which precludes using conventional techniques for cover system construction. These
materials are also prone to large post-construction settlements that must be accounted for in
design.
In general, if the undrained shear strength of the near surface waste is less than about 15 to 20
kPa, the waste may not be able to support a conventional cover system and the bearing capacity
of the waste will be an important consideration in the design process. At undrained shear
strengths below about 10 kPa, waste bearing capacity may become the controlling design
criterion. Guidance on performing foundation bearing capacity analyses can be found in a
number of textbooks, including Lambe and Whitman (1969) and Holtz and Kovacs (1981), and a
number of more focused technical papers and reports, including Bonaparte and Christopher
(1987), Humphrey and Rowe (1991), and Holtz et al. (1995). The three latter references provide
information on the use of geosynthetics reinforcement to increase the bearing capacity of a soft
foundation material (e.g., waste). The critical bearing capacity condition for a cover system over
soft waste will typically occur during construction The critical condition is often associated with
the timeframe during which the leading edge of construction is inducing relatively high shear
stresses in the soft waste. Thus, construction equipment loads must be taken into account.
The engineering evaluation of a cover system over soft waste includes an assessment of overall
bearing capacity, rotational stability, and lateral spreading. These potential failure mechanisms
are illustrated in Figure 6-34. In addition to these stability evaluations, long-term settlement of
DRAFT - DO NOT CITE OR QUOTE
6-60
-------
Direction of Construction
Original Cover System
Deformed
Cover System
Mudwave
(a)
Original Cover System
Rotated Cover System
Soft Waste
Slip Surface
(b)
Original Cover System -
Lateral Spreading
Soft Waste
;«««:<«s&$&gs
(c)
Figure 6-34. Three Potential Failure Mechanisms to be Considered for Cover Systems
Constructed Over Soft Wastes: (a) Bearing Capacity Failure; (b) Rotational
Failure; and (c) Lateral Spreading.
DRAFT - DO NOT CITE OR QUOTE
6-61
-------
the cover system is estimated using classical geotechnical engineering calculation methods for
soil or waste, as appropriate, as described previously in Section 6.4.
The design engineer has several available options when soft waste has inadequate shear strength
to support the overlying cover system. These options include:
• strengthening of the waste by physical solidification;
• strengthening of the waste by preloading;
• strengthening of the waste by dewatering;
• strengthening of the waste by ET drying;
• supporting the cover system over the waste using reinforcement; and
• using lightweight cover system components.
Solidification is defined by EPA as a process in which materials are added to a waste to produce
a solid to achieve one or more goals (Battelle, 1993). In the application being considered herein,
the goals are to increase the waste's shear strength and decrease its compressibility. Agency
guidance on waste solidification technology regulatory status, range of applicability and
limitations, and use on a project-specific basis is given in the agency document (Battelle, 1993).
Typical solidifying agents for this application include cement, cement kiln dust, lime, lime kiln
dust, and fly ash. The final product may vary from a granular, soil-like material to a cohesive
solid, depending on the properties of the solidifying agent and waste and the ratio of the
solidifying agent to waste. One disadvantage of this approach is that the solidification process
causes significant bulking (increase in volume) of the waste. In some cases, this increased
volume can be used to advantage to build up the top elevations of what is initially a flat
impoundment to achieve the sloping grades required for the cover system. As an example of a
solidification project, Bodine and Trevino (1996) describe a case study where a cover system
with a GM/CCL barrier was constructed over an oily sludge storage basin after the sludge was
solidified in-situ using Portland cement. A 3.7-m diameter crane-mounted rotary auger was used
to mix the cement with the sludge.
Strengthening of the waste by preloading involves spreading a layer of soil fill over the waste,
then allowing the waste to consolidate under the weight of the fill. Additional layers of fill can
be placed and the consolidation step repeated. Each consolidation step increases the undrained
shear strength of the waste by about 20 to 25% of the applied vertical stress resulting from the
weight of the fill. Disadvantages of this technique are that it is time consuming, due to the time
required for waste consolidation, it involves multiple construction steps, and it requires
significant amounts of fill. The time duration required for waste consolidation can be decreased
through the use of wick drains. However, installation of wick drains into the soft waste may not
be feasible due to access, settlement, and clogging issues. Vacuum consolidation can be
considered as an alternative means to soil fill for applying a consolidation stress to the soft
waste. However, as with wick drains, installation issues may render this technique unfeasible for
some applications. Guidance on the design of preload systems can be found in Holtz and Kovacs
(1981) and Ladd (1991).
DRAFT - DO NOT CITE OR QUOTE
6-62
-------
Strengthening of the waste mass by dewatering involves the use of drainage trenches or other
means to reduce the water level in the soft waste. A reduced water level has two benefits: (i) as
the water level is pulled down, the effective stress in the waste, and hence the waste's strength, is
increased; and (ii) evaporation from exposed waste above the water table tends to dry out the
waste and increase its shear strength. In some cases, this surface drying can by itself lead to a
stable crust upon which to build a cover system.
Strengthening of the waste by ET drying involves using high moisture uptake plant species to
remove water from the waste by transpiration, thereby strengthening the near-surface waste.
With ET drying, select plant species are planted or hydroseeded over the area to be strengthened.
Because soft waste materials typically have a high moisture content, plants can readily access
moisture in the waste matrix. As plants become established, two complementary benefits occur:
(i) the waste surface dries and a strengthened crust develops; and (ii) the plant roots form a mat
that reinforces and further strengthens the crust. Depending on the type of vegetation selected,
the root mat may extend several inches (as in the case of grasses) or several feet (as in the case of
certain tree species) into the waste. The success of the ET approach is dependent on the physical
properties of the waste and the ability to keep the waste surface dewatered for the period of time
required to establish healthy plant growth. The application of fertilizers or conditioning agents
may be necessary to establish and sustain adequate plant growth. Simple greenhouse testing can
be used to evaluate the potential effectiveness of ET drying. Pilot testing is recommended to
quantify the amount of strengthening that can be achieved in a particular field application.
Geosynthetic reinforcement materials can be used to support cover systems over soft waste. This
technique has found increasing use in recent years. With the approach, one or more layers of
geogrid or GT reinforcement are placed over the soft waste, fill is placed on top of the
reinforcement, and then the cover system is constructed on top of the fill. This technique has
been used successfully with very soft waste materials (i.e., materials with undrained shear
strengths below 10 kPa). Michalski et al. (1995) provide an example of the application of this
technique to the closure of a 10-ha pickle liquor sludge lagoon in Pennsylvania. Closure of the
lagoon required construction of a RCRA Subtitle C cover system, which was supported over the
soft sludge by geogrid reinforcement. Guidance on the design of geosynthetic reinforcement
systems can be found in the technical references cited previously in this subsection, and in
Koerner (1998). While this technique has been shown to be effective, it is not without
limitations. The technique does not reduce the inherent compressibility of the waste. Thus,
when utilized with very soft materials, large total settlements of the cover system may occur.
Also, when using this system over large areas, it may not be possible to build up the required
final grades for the cover system. Even with a saw-tooth final grading plan, fill thicknesses can
be significant to achieve cover system grades of at least several percent. The problem is
exacerbated by the fact that the largest waste settlements will occur at the location where the fill
thickness is greatest, which will tend to reduce cover system grades as the waste settles. The
effects of settlements on cover system grades need to be considered in the design of geosynthetic
reinforcement systems and may also be important for some, or all, of the other technologies
described above.
The final option considered for construction of cover systems over soft waste materials is the use
of lightweight cover system components. Examples of lightweight materials include, from the
bottom of the cover system upward:
DRAFT - DO NOT CITE OR QUOTE
6-63
-------
• geosynthetic reinforcement as a replacement for structural fill;
• lightweight structural fill as a replacement for structural fill; potential lightweight fill
materials include slag, expanded clay and shale, vermiculite, tire chips, and geofoam;
expanded clay and shale materials, manufactured by heating clay or shale in a kiln, are
discussed by Bowders et al. (1997b); the geofoam class of geosynthetics is discussed in
Section 3.7.1 of this guidance document;
• GC drainage layer or a thick needlepunched nonwoven GT as a replacement for a
granular gas transmission layer;
• GCL as a replacement for a CCL; and
• GC drainage layer as a replacement for a granular drainage layer.
An example of a lightweight cover system designed and constructed as part of a CERCLA
remedial action at a soft waste and soil site near Beaumont, Texas, is illustrated in Figure 6-35.
Each of the options for constructing cover systems over soft waste materials have advantages and
disadvantages that must be carefully evaluated for each project application. For most
applications, several of the options will be used in combination to achieve the project design
criteria.
DRAFT - DO NOT CITE OR QUOTE
6-64
-------
0.3 m
0.3 m
.;.:.•;.:.:•'. Su'rfa6e/Prbtect[on Layer :.:••. •.:••" •'.:•
•'.-'•' ;.';.'V-General-Fill Foundation'Layer. •'.-•',. ';•':• •';••
GC Drainage Layer
GM Composite Barrier
GCL
GC GasTransmission Layer
Geosynthetic Reinforcement (where needed)
Regraded Existing Fill
Figure 6-35. Example of a Lightweight Cover System Constructed Over Soft Waste and
Soil at CERCLA site near Beaumont, Texas.
DRAFT - DO NOT CITE OR QUOTE
6-65
-------
Chapter 7
Lessons Learned
7.1 Introduction
As discussed in Section 1.6.1 of this guidance document, there have been a number of
documented cases where cover systems at waste management sites have not performed as
intended. The primary factors contributing to the cover system problems in most cases were
inadequate design and construction. Many of these problems occurred during, or shortly after,
construction. Several, however, did not occur until one or more years after the completion of
construction. The costs of remedying cover system problems can be significant, especially if the
problems involve slope instability, or if they impact maintenance and can recur (e.g., excessive
erosion). Daniel and Gross (1996) summarized the mechanisms that can adversely affect the
performance of each component of a cover system. These mechanisms factors are presented in
Table 7-1.
Table 7-1. Mechanisms that can adversely affect cover system performance (modified
from Daniel and Gross, 1996).
Layer
Surface Layer
Protection Layer
Drainage Layer
Barrier
Gas Collection Layer
Foundation Layer
Factor
Insufficient or excessive slope
Erosion by water and/or air
Slope instability
Insufficient nutrients or inadequate soil texture to support vegetation
Inadequate soil thickness and thus water storage capacity to maintain
adequate vegetation
Undesirable vegetative species
Erosion by water
Slope instability
Accidental human intrusion
Intrusion by burrowing animals
Root penetration
Inadequate soil texture to support vegetation
Excessive clogging
Insufficient flow rate capacity
Insufficient number or flow rate capacity of outlets
Freeze effects
Slope instability
Cracking due to wet-dry effects, freeze-thaw effects, differential
settlements, seismic motions
Deep root penetration
Insufficient resistance to gas flow
Slope instability
Creep of all materials (CCL, GCL, GM, asphalt)
Insufficient coverage over waste
Insufficient flow rate capacity
Insufficient strength
DRAFT - DO NOT CITE OR QUOTE
7-1
-------
The purpose of the remainder of this section is to share recent information on experiences and
lessons learned related to the design and construction of cover systems in a variety of situations.
These experiences and lessons learned have been organized by the following subject areas:
• soil barriers;
• GM barriers;
• slope stability;
• waste settlement;
• stormwater management and erosion control;
• gas pressures; and
• miscellaneous.
Consistent with the discussion in Section 1.6.1 of this document, EPA believes that improvement
can, and should, be made in the design and construction of cover systems. The information
presented in this chapter is intended to alert engineers to past problems in the design and
construction of these systems. By application of the lessons learned from this chapter, future
design and construction can be improved and potential problems can be prevented.
7.2 Soil Barriers
Experiences and lessons learned with respect to the hydraulic performance of soil barriers in
cover systems have focused primarily on the use of CCLs in this application. Bonaparte et al.
(2002) discussed available case studies on CCLs and GCLs, from which the majority of the
following information has been extracted.
7.2.1 Test Plots in Omega Hills, Wisconsin
One of the first detailed field studies on the performance of CCLs in landfill cover systems was
described by Montgomery and Parsons (1989, 1990). Three large test plots with different cover
system designs were constructed on top of the closed Omega Hills landfill and monitored for
four years. The purpose of the field study was to compare the performance of the different cover
systems. The landfill had accepted MSW and is located approximately 30 km northwest of
Milwaukee, Wisconsin.
The cross sections of the three test plots are shown in Figure 7-1. Test plot 1, consisting of a
0.15-m thick topsoil layer overlying a 1.2-m thick CCL barrier, was representative of the existing
cover system on part of the landfill. Test plot 2 involved the same thickness of CCL, but a
thicker (i.e., 0.45-m thick) topsoil layer intended to promote better vegetative growth and
thereby enhanced ET. Test plot 3 incorporated a layer of coarse-grained soil (sand) interbedded
between two CCLs. The concept for the third plot was to take advantage of the capillary barrier
effect (see Section 1.1.2), with the sand layer promoting retention of water in the upper CCL and
enhanced ET. All test plots were constructed on the landfill's 3H: IV sideslopes. The CCL
material was classified as CL according to the Unified Soil Classification System (USCS) and
had a high silt content. The soil was placed and compacted in 0.15-m thick lifts to a hydraulic
conductivity no greater than 1 x 10"9 m/s, based on laboratory hydraulic conductivity tests on
DRAFT - DO NOT CITE OR QUOTE
7-2
-------
Test Plot 1
Test Plot 2
Test Plot 3
1
0.15m
1.2m
0.45m
1.2m
0.15m
0.6 m
0.3m
0.6m
CCL
,:Sand;
CCL
Figure 7-1. Cross Sections of Cover System Test Plots at a Landfill in Omega Hills,
Wisconsin (modified from Montgomery and Parsons, 1989). Test Plot 1 is
Representative of the Existing Landfill Cover System.
"undisturbed" small-diameter samples of the compacted soil. The sand in test plot 3 was a clean,
washed, medium sand. The topsoil consisted of uncompacted clay loam to silty clay loam and
was seeded with a mixture of grasses.
The test plots contained two principal data collection systems. The first system was a collection
lysimeter installed beneath the test plot to collect water that percolated through the cover soils
and allow quantification of the rate of percolation. The lysimeter consisted of, from top to
bottom, a GT filter, a GC drainage layer, and a GM. The second data collection system was
designed to collect and measure surface runoff.
The test plots were constructed from September 1985 to July 1986, and data collection and
analysis began in August 1986. Measurements were obtained of precipitation, runoff,
percolation, and other parameters such as temperature. Soil moisture content was monitored
using neutron probes, and, until September 1988, soil matric potential was monitored using
tensiometers.
The weather during the 12-month period from September 1986 through August 1987 was near
normal. The period of September 1987 through August 1988 was dominated by a drought,
which occurred during May through August. These months were characterized by substantially
below average rainfall and temperatures that averaged 6 °C above normal. The drought reduced
the cover vegetation to a dry, dormant state and caused cracking of the cover soils. The third
year of data collection (September 1988 to August 1989) saw a return to normal conditions and a
reduction in surface cracking. The nine-month period from September 1989 through April 1990
included a dry fall, a mild winter, and a spring with normal precipitation, but erratic temperature
DRAFT - DO NOT CITE OR QUOTE
7-3
-------
fluctuations. At the end of this monitoring period, cover vegetation was vigorous and included a
number of plant species not in the original seed mix.
A summary of data through April 1990 is presented in Table 7-2. The key parameter is the
quantity of percolation, i.e., flow rate of water into the lysimeter. In test plots 1 and 2, the
percolation during the first year was 2 and 7 mm/year (6 x 10"11 to 2 x 10"10 m/s), respectively.
However, by the third year, these values had increased to 56 and 98 mm/year (2 x 10"9 and
3 x 10"9 m/s), respectively. For test plot 3, which was designed with the intention of maintaining
moisture in the upper CCL, the percolation rate remained more consistent and was found to
range from 22 to 41 mm/year (7 x 10"10 to 1.3 x 10"9 m/s) during the first three years. In
September 1988, at the end of the third year, 2-m deep test pits were excavated in each test plot,
outside the area of the lysimeters. Examination of the test pits revealed that the CCLs in the test
plots were in a similar condition:
• the upper 0.20 to 0.25 m of the CCLs were weathered and blocky;
• cracks 6 to 12 mm wide extended about 0.9 to 1 m into the CCLs in test plots 1 and 2 and
through the entire thickness of the uppermost CCL in test plot 3;
• the base of the CCLs in test plots 1 and 2 appeared to be undamaged;
• roots penetrated 0.20 to 0.25 m into the CCLs in a continuous manner, and some roots
extended as deep as 0.75 m into cracks in the CCLs; and
• the moisture contents in the upper portion of the CCLs were near the shrinkage limit.
The drought conditions in the second year of the study period apparently caused desiccation of
the CCL in test plots 1 and 2, which led to a significantly increased CCL hydraulic conductivity
in subsequent years. Although the CCLs in these tests plots may have initially had a hydraulic
conductivity of 1 x 10"9 m/s or less, the desiccation damage caused the CCLs to no longer have
this low level of hydraulic conductivity.
Table 7-2. Summary of performance-related information for field test plots at Omega
Hills Landfill (data from Montgomery and Parsons, 1990).
Test Plot Year Precipitation Runoff Percolation
Designation (mm) (mm) (mm)
1 1986-87 896 180 2
1987-88 578 38 5
1988-89 823 56 56
1989-90 350 44 33
2 1986-87 896 109 7
1987-88 579 38 30
1988-89 823 51 98
1989-90 350 22 31
3 1986-87 896 97 40
1987-88 579 38 22
1988-89 823 66 41
1989-90 350 23 16
DRAFT - DO NOT CITE OR QUOTE
7-4
-------
In May 1990, a second test pit was excavated in test plot 1. No major cracks were observed in
the CCL, in contrast to the pronounced cracking of the upper portion of the CCL observed in the
September 1988 test pits. The CCL appeared uniformly moist, probably as a result of spring
precipitation. Roots did not appear to be deeper or more dense than observed in the earlier test
pits. The base of the CCL still appeared to be homogeneous, moist, and intact. It is noteworthy
that while the physical condition of the CCL in test plot 1 appeared to have improved,
percolation through the CCL in 1990 remained at a high level.
For test plot 3, cracking of the uppermost CCL allowed significant amounts of water to enter the
sand drainage layer. Discharge of water from the sand layer was found to occur within hours of
the start of precipitation events, suggesting rapid transmission of water through the upper CCL
due to preferential flow through cracks. Moisture in the sand drainage layer probably helped to
protect the underlying CCL from damage. The capillary barrier in test plot 3 did not function as
well as anticipated. It was expected that the sand drainage layer would help the overlying CCL
retain moisture, but the uppermost CCL quickly cracked.
As of April 1990, percolation through test plots 1 and 2 was approximately 9% of precipitation,
and percolation through test plot 3 was approximately 4.6% of precipitation.
The principal lessons learned from the Omega Hills study are that in a fairly short period of time
(3 years), CCLs overlain by only 0.15 to 0.45 m of topsoil are subject to desiccation, cracking,
and increases in hydraulic conductivity. The CCLs were incapable of "surviving" under these
conditions with a hydraulic conductivity of 1 x 10"9 m/s or less.
7.2.2 Test Plots in Kettleman City, California
Corser and Cranston (1991) and Corser et al. (1992) described three test plots constructed at a
landfill in Kettleman City, California. Cross sections of the test plots are shown in Figure 7-2.
Test plot 1 consisted of a 0.9-m thick CCL overlain by an exposed 1.5-mm thick HDPE GM.
Test plot 2 consisted of the same profile as test plot 1, except that 0.6 m of topsoil covered the
GM. For test plot 3, 0.45 m of topsoil covered the CCL and no GM was present. A portion of
the test plots was flat, and another sloped at 3H: IV. The test plots were constructed to study the
factors that influence desiccation of CCLs used in cover systems.
The CCL material was a high plasticity clay that the site owner intended to use in cover system
construction for approximately 30 ha of the landfill. The clay had an average liquid limit of 66%
and plasticity index of 48%. Instrumentation for the test plots consisted of thermistors to
monitor temperature in the topsoil and CCL and tensiometers to measure soil water potential.
Corser and Cranston (1991) summarized the first six months of data collection. At the end of the
six-month period, the surfaces of the CCLs were exposed over an area of 1.5 m x 1.5 m to
observe and document cracking patterns.
Test plot 1 did not represent a cover system design but, instead, an exposed HDPE GM/CCL
composite liner during the construction or operations phase. The clay exhibited some drying and
cracking in areas where the GM was not in contact with the CCL. In other areas, where the GM
was in contact with the CCL, the moisture content of the CCL at the surface had increased. It
appears that the high temperature of the exposed HDPE GM caused heating and drying of the
DRAFT - DO NOT CITE OR QUOTE
7-5
-------
underlying CCL. In some areas (e.g., around wrinkles in the GM), moisture could migrate away
via vapor transport. In other areas, the moisture could condense during cooler periods, causing
moistening of the soil. In any case, there clearly was desiccation of the CCL beneath some
portions of the exposed GM.
Test Plot 1
1.5 mm HOPE Gl\
0.9m
Test Plot 2
Test Plot 3
0.6 m
0.9m
Topsoil
0.45m
-1.5 mm HDPEGM
0.9m
Topsoil
Figure 7-2. Cross Sections of Cover System Test Plots at a Landfill in Kettleman City,
California (modified from Corser and Cranston, 1991).
Test plot 3 did not perform well during the summer season. The CCL dried, and cracking was
observed at its surface. In contrast, test plot 2 performed well. There was no evidence of drying
or cracking of the CCL.
Although the test plots were observed for only six months, significant deterioration of the CCLs
was observed in test plots 1 and 3. Only test plot 2, in which the CCL was covered with a GM
and 0.6 m of topsoil, performed well. The observations from Kettleman City are consistent with
those of Omega Hills and suggest that perhaps the only practical way to protect a CCL from
desiccation damage in typical cover system applications is to cover it with a GM overlain by a
sufficiently thick layer of soil.
7.2.3 Cover Systems in Maine
The Maine Bureau of Remediation and Waste Management (1997) reported the results of
laboratory and field hydraulic conductivity measurements for four CCL barriers in actual MSW
landfill cover system applications. The laboratory tests were conducted on "undisturbed" small-
diameter samples collected from the constructed CCLs. It appears that all four cover systems
were installed using methods of construction and CQA practices that are representative of
landfill industry practices presently used in the U.S.
Cumberland Site. The Cumberland MSW landfill, a 2 ha facility, was closed in 1992 with a
cover system consisting of a 0.15-m thick vegetated topsoil layer underlain by a 0.45-m thick
silty clay CCL. Underlying the CCL are sand-filled trenches that serve to collect and convey
landfill gas. Laboratory hydraulic conductivity tests were performed on CCL samples during
DRAFT - DO NOT CITE OR QUOTE
7-6
-------
construction and in a post-construction investigation program conducted in 1994. A sealed
double-ring infiltrometer (SDRI) test was also performed in 1994.
At the time of construction, the average CCL hydraulic conductivity measured in the laboratory
was 5 x 10"10 m/s. In the 1994 investigation, the laboratory-measured hydraulic conductivity had
increased to 1 to 2 x 10"9 m/s. The field hydraulic conductivity, measured with the SDRI in
1994, was 6 x 10"8 m/s. It is not certain whether the CCL originally had a field hydraulic
conductivity greater than 1 x 10"9 m/s since field testing was not performed at the time of
construction.
Vassalboro Site. The Vassalboro MSW landfill occupies 11.6 ha and was closed in 1990 with a
cover system consisting of, from top to bottom: a 0.15-m thick sludge-amended topsoil layer; a
0.45-m thick glacial till CCL; and a gas collection layer. Laboratory hydraulic conductivity tests
were performed at the time of construction, and again in 1994. An SDRI test was also performed
in 1994.
The average hydraulic conductivity of the CCL measured in the laboratory at the time of
construction was 2 x 10"9 m/s. In 1994, the laboratory-measured hydraulic conductivity values
ranged from 9 x 10"9 to 5 x 10"8 m/s and the field-measured hydraulic conductivity was
2 x 10"8 m/s. It appears that the hydraulic conductivity of the CCL increased by about an order
of magnitude from 1990 to 1994.
Yarmouth Site. The Yarmouth MSW landfill, a 2.5 ha facility, was closed in 1990 with a cover
system consisting of, from top to bottom: a 0.15-m thick sludge-amended topsoil layer; a 0.45-m
thick silty clay CCL; and a gas collection layer. Laboratory hydraulic conductivity tests were
performed at the time of construction, and again in 1994 and 1996. An SDRI test was also
performed in 1994 and 1996.
Laboratory hydraulic conductivity tests conducted in 1990 indicated an average CCL hydraulic
conductivity of 8 x 10"10 m/s. In a 1994 investigation, the average measured laboratory hydraulic
conductivity was 3 x 10"9 m/s, and, in 1996, the laboratory-measured hydraulic conductivity was
in the range of 2 x 10"8 to 2 x 10"7 m/s, or about 20 to 100 times larger than in 1990. The field-
measured hydraulic conductivity was 2 x 10"9 m/s in 1994 and 2 x 10"8 m/s in 1996. There is a
clear trend of increasing hydraulic conductivity over time, with the magnitude of increase being
one to two orders of magnitude over the six-year study period.
Wai dob oro Site. The Waldoboro MSW landfill encompasses 1.6 ha and was closed in 1991 with
a cover system consisting of, from top to bottom: a 0.15-m thick sludge-amended topsoil layer; a
0.45-m thick silty clay CCL; and a gas collection layer. Laboratory hydraulic conductivity tests
were performed at the time of construction, and again in 1993 and 1996. An SDRI test was also
performed in 1993 and 1996.
Laboratory hydraulic conductivity tests indicated that the CCL hydraulic conductivity increased
over time from an initial average value of about 5 x 10"10 m/s (1991) to 1 x 10"8 m/s (1993) and
then to 3 x 10"8 m/s (1996). The field hydraulic conductivities were 1 x 10"8 m/s (1993) and
DRAFT - DO NOT CITE OR QUOTE
7-7
-------
4 x 10"8 m/s (1996). Thus, the data indicates that the hydraulic conductivity increased by about
two orders of magnitude over a five-year period.
Discussion. The observations from these four cover system case studies are consistent with
those of the other sites mentioned previously in this chapter. All of the available field
performance data indicate that a CCL barrier overlain by a relatively thin layer of topsoil or
protection soil (0.15 to 0.45 m thick), and without a GM above the CCL, cannot maintain a
hydraulic conductivity of 1 x 10"9 m/s or less. From analysis of the condition of the four CCL
barriers at these sites, it appears that desiccation was the most significant factor leading to an
increase in field hydraulic conductivity. Freeze/thaw may also have contributed to the observed
degradation in CCL performance. Penetration of plant roots into the CCL was also observed.
7.2.4 Test Plots in Live Oak, Georgia and Wenatchee, Washington
Lane et al. (1992), Khire (1995), and Khire et al. (1997, 1999) reported on field water balance
studies for three 30 m x 30 m cover system test plots at two landfills, one near Atlanta, Georgia
("Live Oak") and the other near East Wenatchee, Washington ("Wenatchee"). The sites were
selected to represent humid and semi-arid climates, respectively. The Live Oak test plot has a
cover system with a 0.9-m thick CCL overlain by a 0.15-m thick silty topsoil layer. In
Wenatchee, one test plot has the same cover system as at the Live Oak site except that the CCL
is 0.6 m thick, and the other test plot models a capillary barrier consisting of a 0.75-m thick layer
of uniformly-graded medium sand overlain by a 0.15-m thick silt topsoil layer. Climate, runoff,
percolation, and soil moisture data collected between 1992 and 1995 were reported by Khire
(1995) and Khire et al. (1997, 1999), and data collection is still ongoing as of 2001. Details of
the water balance analyses for these test plots are provided in Chapter 4 of this guidance
document. Importantly, the results of these field studies show nearly 250 mm of percolation
through the Live Oak test plot in a period of 21/2 years. Percolation through the CCL barrier test
plot at the Wenatchee site over a roughly similar period was much less than at Live Oak, due to
the more arid site conditions, but still significant (more than 30 mm). Percolation through the
capillary barrier test plot was low, only about 5 mm. The conclusions for these data are
consistent with those presented previously in this chapter. Percolation rates through
inadequately-protected CCL barriers are relatively high. The limited results for the capillary
barrier at the Wenatchee site are encouraging.
7.2.5 Test Plots in Hamburg, Germany
Melchior et al. (1994) and Melchior (1997a,b) described what may be the most extensive test
plot program to date involving CCLs for cover systems. Test plots with four different cover
system cross sections, shown in Figure 7-3, were constructed over a MSW landfill in Hamburg,
Germany. The test plots with CCLs were constructed in 1987, and the test plots with GCLs were
constructed in 1995. Each test plot is 10 m wide and 50 m long and is located on the relatively
flat (i.e., 4% slope) top deck or on the 5H: IV sideslopes of the landfill. Climate, lateral
drainage, runoff, percolation, soil moisture content, and soil water potential data are being
collected.
The CCL material at the Hamburg site consisted of a glacial till comprising 17% clay, 26% silt,
52% sand, and 5% gravel. The principal clay minerals in the clay-sized fraction were (in
decreasing abundance) illite, smectite, and kaolinite. The soil liquid limit was 20%, and the
DRAFT - DO NOT CITE OR QUOTE
7-8
-------
Test Plot S1/F1
Test Plot S2/F2/F3
Test Plot S3
Test Plot B1/B2
0.75m
0.25m
0.6 m
0.2m
Topsoil
Sand Drainage
'•v1./Layer',V::.--'
0.75m
0.25m
GM Lysimeter-
0.75m
0.25m
2mm HDPEG
0.6 m
0.2m !
GM Lysimeter
0.4 m
0.6m
0.25m
'Sand prainac|fe
''
Topsoil
'•'•' Sand '•;'•.'.'.'•
0.3 m
0.15m
Topsoil
Sarid Drainagte
GCL-
GM Lysimeter-
GM Lysimeter—'
Figure 7-3. Cross Sections of Cover System Test Plots at a Landfill in Hamburg,
Germany (modified from Melchior et al., 1994; Melchior, 1997a).
plasticity index was 9%. The soil was placed in 0.20-m thick compacted lifts at two percentage
points wet of the standard Proctor optimum moisture content and to an average degree of
compaction of 96% of the standard Proctor maximum dry density. The geometric mean
hydraulic conductivity of the CCLs was 2.4 x 10"10 m/s, based on laboratory hydraulic
conductivity tests on "undisturbed" small-diameter samples of the compacted soil. The CCL
material at the Hamburg site was significantly different from that at the Omega Hills and
Kettleman City sites. At Omega Hills, the CCL material was a low-plasticity clay (CL)
containing a large amount of silt, which can make the CCL vulnerable to shrinkage cracking.
The Kettleman City CCL material was a high-plasticity clay (CH). At Hamburg, the CCL
material contained more than 50% sand- and gravel-sized particles and would therefore be
classified as a clayey sand (SC). Clayey sands tend to be less vulnerable to shrinkage cracking
than clays (especially highly plastic clays) that contain relatively few coarse-grained particles.
Percolation rates through the CCLs from 1988 to 1995 for the test plots with a 4% slope (test
plots Fl, F2, and F3) are summarized in Table 7-3. Percolation rates through the CCLs (i.e.,
drainage from the underlying lysimeters) from 1988 to 1995 for the test plots with a 20% slope
(test plots SI, S2, and S3) are summarized in Table 7-4. Also shown in Tables 7-3 and 7-4 are
the lateral flow rates from the sand drainage layers that overlie the CCLs. The last column in the
tables expresses percolation through the CCLs as a percentage of the lateral flow from the sand
drainage layers.
DRAFT - DO NOT CITE OR QUOTE
7-9
-------
Table 7-3. Summary of field performance data for Hamburg, Germany test plots
containing CCLs and at 4% slope (data from Melchior, 1997a).
Percolation/
Test Plot Year Lateral Drainage Percolation Drainage
Designation (mm) (mm) (%)
F1 1988 368 7 2
1989 183 8 4
1990 286 18 6
1991 187 9 5
1992 226 103 46
1993 253 174 69
1994 247 166 67
1995 156 164 105
F2 1988 293 3.5 1
1989 156 0.6 0.4
1990 263 0.4 0.1
1991 171 0.5 0.3
1992 313 0.8 0.3
1993 412 1.3 0.3
1994 409 1.8 0.4
1995 310 1.7 0.5
F3 1988 367 4.1 1.1
1989 155 1.4 0.9
1990 262 2.6 1.0
F3 1991 168 2.0 1.2
1992 326 3.5 1.1
1993 481 5.0 1.0
1994 431 5.2 1.2
1995 328 5.2 1.6
As can be observed from inspection of the data in Tables 7-3 and 7-4, test plots Fl, SI and S3,
which did not have a GM overlying the CCL, underwent large increases in percolation rate
within three to four years after installation. In particular, the summer of 1992 was very dry in
Hamburg, and the subsequent fall season was very wet. By 1992, actual percolation rates
exceeded 100 mm/year in two of the three CCL test plots. The third CCL test plot exceeded this
percolation value by 1993. Excavations made in 1993 confirmed that the CCLs in these test
plots were cracked. Barely visible fissures were observed between soil aggregates (around 50
mm in diameter). By 1995, plant roots were observed to have extended more than 1 m into the
cover system, reaching the upper parts of the CCLs. In summary, the performance of the test
plots containing a CCL without GM protection has been poor. The apparent problem is gradual
deterioration of the CCLs caused by desiccation during a particularly dry summer. Detailed
results for test plot SI are presented in Figure 7-4 for illustration purposes.
Percolation rates through test plots S2, F2, and F3, which contain a CCL overlain by a GM, have
been very low. Test plots F2 and S3, which incorporate a continuously-welded HDPE GM, had
average measured percolation rates of 1.3 mm/year, while test plot F3, which has an overlapped
(not welded) GM, exhibited an average measured percolation rate of 3.6 mm/year. Melchior
(1997a) indicated that the measured percolation is primarily due to thermally-driven unsaturated
flow of pore water in the CCL, not to leakage through the GM.
DRAFT - DO NOT CITE OR QUOTE
7-10
-------
Table 7-4. Summary of field performance data for Hamburg, Germany test plots
containing CCLs and at 20% slope (data from Melchior, 1997a).
Percolation/
Test Plot Year Lateral Drainage Percolation Drainage
Designation (mm) (mm) (%)
S1 1988 386 1.9 0.5
1989 247 3.1 1.2
1990 318 13 4
1991 177 13 7
1992 289 48 17
1993 343 136 40
1994 344 150 44
1995 229 150 66
S2 1988 355 0.6 0.2
1989 237 0.3 0.1
1990 321 0.5 0.2
1991 192 0.7 0.4
1992 330 1.0 0.3
1993 390 1.7 0.4
1994 389 3.0 0.8
1995 297 2.8 0.9
S3 1988 396 84 2
1989 234 14 6
1990 319 31 10
1991 200 3 16
1992 279 117 42
1993 263 171 65
1994 248 184 74
1995 151 201 133
The two test plots (Bl and B2) containing GT-encased GCLs were constructed in early 1995
with an 8% slope. The GCLs were covered with a 0.15-m thick sand drainage layer and a 0.3-m
thick topsoil layer. Melchior (1997a) reported that both GCL test plots performed very well
through the first winter after installation. However, after a dry summer (1995), significant
percolation occurred through both GCLs. Through four months in the fall of 1995, percolation
through the two test plots was 45 and 63 mm. Melchior reported that during the 1995/1996
winter, the GCLs did not rehydrate and swell enough to completely heal the preferential flow
paths caused by the previous summer's desiccation. In part, this may be due to calcium for
sodium ion exchange within the bentonite.
With respect to the CCL test plots, the results from the Hamburg test site are consistent with
those from the Omega Hills and Kettleman City test sites, even though the CCL materials for the
three sites were different. The up to 0.75 m of topsoil placed over the CCLs at the different sites
was not sufficient to maintain the low hydraulic conductivity of the CCLs. It appears that a CCL
placed in a cover system without a GM and a sufficient thickness of soil covering the GM is
likely to fail to maintain a hydraulic conductivity < 1 x 10"9 m/s, at least for the considered sites
and surface/protection soil thicknesses. It is emphasized that from a practical perspective, if the
DRAFT - DO NOT CITE OR QUOTE
7-11
-------
E
25-
20 -=
15-
E 3
Lateral Drainage Above CCL
Percolation Through CCL.
1 I I I I I i I I I I r
1988 1989 1990 1991
i i i r
1992 1993 1994
1995
Figure 7-4. Summary of Field Data for Test Plot S1 with a CCL Barrier at a Landfill In
Hamburg, Germany (modified from Melchior, 1997a).
CCL is to have a chance of maintaining a low hydraulic conductivity for an extended period, the
CCL must be protected with both a GM and a sufficiently thick layer of cover soil above the
GM. Furthermore, if a GCL is used in lieu of a CCL, the GCL must be chemically-compatible
with adjacent soils.
7.2.6 Test Plots in Albuquerque, New Mexico
Dwyer (1997, 1998, 2001) described the U.S. Department of Energy (DOE) funded Alternative
Landfill Cover Demonstration (ALCD) project, which involved the construction and monitoring
of six test plots with different cover system configurations at the Kirtland Air Force Base in
Albuquerque, New Mexico. The six cover system types being evaluated are shown in Figure
7-5. To provide good vegetation coverage during the growing season, the plots were seeded with
a mixture of warm season and cold season native grasses.
The test plots were constructed in 1995 and 1996. Each test plot is 13 m wide by 100 m long,
crowned in the middle, and sloped at 5% in both length directions from the crown. For each test
plot, one slope ("western slope") is monitored under the ambient conditions existing at the site.
The other slope ("eastern slope") has a sprinkler system to provide a hydrologic stress to the
cover systems. Continuous water balance and meteorological data are being collected for the test
plots. The plots are heavily instrumented to quantify measurable water balance variables
(precipitation, surface runoff, lateral drainage, percolation, and soil moisture changes.)
Instrumentation includes collection lysimeters to monitor percolation and time domain
DRAFT - DO NOT CITE OR QUOTE
7-12
-------
Hydraulic
I Barrier
Hydraulic
Barrier
^
Alternative Test Plot 1
0.15
0.45m
Alternative Test Plot 3
0.15m
Alternative Test Plot 4
0.15m
—| Hydraulic
J Barrier
Sand Filter
Layer
Capillary
Barrier
Anisotropic
Capillary
Barrier
Figure 7-5. Cross Sections of Cover System Test Plots at Kirtland Air Force Base in
Albuquerque, New Mexico (modified from Dwyer, 1997).
DRAFT - DO NOT CITE OR QUOTE
7-13
-------
reflectometry (TDR) moisture sensors to monitor the soil water content within the cover system.
Each test plot is briefly described below, and summary percolation data is presented in Table 7-
5.
Test plot 1 has a RCRA "Subtitle D" prescriptive minimum criteria cover system (hydraulic
barrier type). The hydraulic barrier for this system consists of a 0.45-m thick CCL (k < 1 x 10"7
m/s). The measured annual percolation through this cover system during the first three years of
monitoring averaged 4.82 mm. Dwyer (1998) reported "As expected, the subtitle D soil cover
performed poorly.... Desiccation cracking, freeze/thaw cycles, root penetration, and earthworm
and insect activity have acted to increase the permeability."
Test plot 2 has a RCRA "Subtitle C" equivalent minimum technology guidance cover system
(hydraulic barrier type). The hydraulic barrier consists of a 0.6-m thick CCL (k < 1 x 10"9 m/s)
overlain by a 1-mm thick LLDPE GM. Importantly, the GM had eight 1 cm2 holes cut into it to
simulate installation-induced defects. Reported average annual percolation is 0.13 mm. Dwyer
(1998) reported: "The other baseline cover - the subtitle C compacted clay cover - had little
percolation for most of the year. However, in the past few months percolation has been evident,
and the percolation rate is expected to slightly increase with time. One problem with this system
is that the geomembrane hampers the ability of the barrier layer to dry by ET; consequently, as
additional moisture infiltrates the barrier layer it eventually creates percolation. " With respect
to this cover system, two additional comments are provided: (i) the frequency and size of holes
in the GM component of the cover system are significantly larger than would normally be
anticipated in a good GM installation; and (ii) evaluation by the researchers involved in the
project indicate that percolation through the CCL may be primarily due to concentrated flow
through desiccation cracks that developed during construction.
Alternative test plot 1 is identical to test plot 2 except that a GCL is used in lieu of a CCL.
Reported average annual percolation is 1.81 mm. Dwyer reported (1998): "The GCL cover is
not performing as well as expected. There are eight 1 cm2 defects in the geomembrane. It is
hypothesized that moisture moved through the geomembrane via defects or diffusion and
penetrated the GCL seams prior to the seams 'full hydration and sealing. The GCL could also
have been damaged during construction, despite very tight quality control, or through root
intrusion.'" Dwyer (2001) gave two additional hypotheses for the apparent increase in GCL
permeability: (i) "the bentonite within the geosynthetic clay liner has desiccated and does not
fully repair itself after rewetting"; and (ii) "the soils in the Southwest or dry environments are
susceptible to ion exchange problems that increase the permeability of the liner".
Alternative test plot 2 has a capillary barrier type of cover system. The cover system has the
following layers, from the surface down: (i) 0.3-m thick topsoil layer; (ii) 0.15-m thick graded
sand filter layer; (iii) 0.22-m thick gravel drainage layer; (iv) 0.45-m thick compacted finer-
grained soil component of the capillary barrier; and (v) 0.3-m thick sand coarser-grained soil
component of the capillary barrier. Reported average annual percolation is 0.87 mm. Dwyer
(1998) reported: "The capillary barrier cover also showed a higher than expected percolation
rate for the first year, but the rate is slowing significantly as the surface vegetation thickens with
native grasses and shrubs."
DRAFT - DO NOT CITE OR QUOTE
7-14
-------
Alternative test plot 3 includes an "anisotropic" capillary barrier, which is a type of capillary
barrier intended to promote unsaturated lateral movement of water through certain soil layers.
The components of this system are, from the surface down: (i) 0.15-m thick surface layer
consisting of 75% local topsoil and 25% pea gravel; (ii) 0.6-m thick finer-grained soil
component of the capillary barrier; (iii) 0.15-m thick fine sand interface layer (wieking layer)
intended to promote lateral drainage under unsaturated flow conditions; and (iv) 0.15-m thick
pea gravel coarser-grained soil component of the capillary barrier. Reported average annual
percolation is 0.16 mm. Dwyer (1998) reported: "The anisotropic barrier andETcover are
both performing very well. Their percolation rates have decreased, as with the capillary barrier,
through increased transpiration from the vegetation growth. Recently, the percolation rates of
both of these covers have fallen below that of the compacted clay cover."
Alternative test plot 4 is an ET barrier type of cover system consisting of, from the surface down:
(i) 0.15-m thick topsoil layer; and (ii) 0.9-m thick native soil layer. Reported average annual
percolation is 0.19 mm. Dwyer (2001) reported: "The evapotranspiration cover appears to be
leading the way in the third year of testing. This test reveals that in dry environments a well-
designed simple soil cover is not only the cheapest alternative but also the most effective at
controlling infiltration"
Table 7-5. Summary of field performance data for Albuquerque, New Mexico test plots
(data from Dwyer, 2001).
Qrt »-^<"k l*^ + i <"*»•» /rvt rvt \
Year
1997 (May to Dec)
1998
1999
2000 (Jan to June)
Average
The ALCD project will provide additional valuable information as it is monitored for a period of
at least five years. Already, the inadequacy of the "Subtitle D" minimum technology guidance
cover system has been demonstrated and the effectiveness of the "Subtitle C" equivalent
minimum technology guidance cover system is being confirmed. Percolation results to date for
the test plots with the anisotropic capillary barrier and ET barrier are also promising. To date,
data provided from this demonstration has been favorably considered by regulators to allow for
the use of alternative cover systems in lieu of a prescriptive cover in several areas in the
southwestern United States.
7.2.6 Test Plots in Los Alamos, New Mexico
Nyhan et al. (1997) described the performance of sixteen test plots constructed at Los Alamos
National Laboratory for the Protective Barrier Landfill Cover Demonstration. The plots had four
different cover system configurations, which were each constructed on slopes of 5, 10, 15, and
20%. None of the plots was vegetated, apparently to simulate conditions in which plants
DRAFT - DO NOT CITE OR QUOTE
7-15
Precipitation
Collected
(L)
154,585
169,048
130,400
28,151
Test
Plotl
10.62
4.96
3.12
0.00
4.82
Test
Plot 2
0.12
0.30
0.04
0.00
0.13
r ci \* \siatr
Alt. Test
Plotl
1.51
0.38
4.31
0.00
1.81
\Jt i yi i ii i tf
Alt. Test
Plot 2
1.62
0.82
0.85
0.00
0.87
Alt. Test
Plots
0.15
0.14
0.28
0.00
0.16
Alt. Test
Plot 4
0.22
0.44
0.01
0.00
0.19
-------
provided no transpiration. Precipitation, runoff, lateral drainage, percolation, and soil water
content are being measured for each test plot.
The four cover system cross sections that were constructed are as follows:
• Test cover 1: the "conventional Los Alamos design" with, from top to bottom, 0.15 m of
loam topsoil, 0.76 m of silty sand, and 0.3 m of gravel.
• Test cover 2: the "EPA design" with, from top to bottom, 0.15 m of loam topsoil, a GT
filter/separator, 0.3 m of drainage sand, and a 0.6-m thick bentonite clay-sand CCL.
• Test cover 3: the "loam capillary barrier design" with 0.6 m of loam topsoil overlying
0.76 m of fine sand.
• Test cover 4: the "clay loam capillary barrier design" with 0.6 m of clay loam topsoil
overlying 0.76 m of fine sand.
Test cover performance data presented by Nyhan et al. (1997) for the first 41/2 years of
monitoring show that Test cover 2 has performed better than the other cover system
configurations. There has been no evidence of percolation for test cover 2 even though its CCL
was only protected by 0.45 m of soil. The bentonite clay mixed in with sand to form the
hydraulic barrier apparently helped the water balance at the site (Bonaparte et al., 2002). The
highest amount of percolation was recorded for test cover 1; measured percolation rates for the
test cover 1 plots ranged from 174 mm for the 5% slope to 31 mm for the 25% slopes over the
41/2 -year monitoring period. Measured percolation rates for test covers 3 and 4, respectively,
ranged from 76 and 48 mm for the 5% slope, 36 and 0 mm for the 10% slope, and 0 mm for both
cover system cross sections on the 15% and 25% slopes.
Even though test cover 2 appears to have a favorable water balance, there is still the concern that
the CCL may degrade over time. Based on the other field studies discussed in this section,
desiccation of CCL barriers in cover systems is a distinct long-term possibility.
7.3 GM Barriers
7.3.1 Percolation through GM Barriers
Several of the soil barrier studies described in Section 7.2 included test plots containing GMs.
The studies of Melchior et al. (1994) and Melchior (1997a,b) provide very good results for cover
system test plots containing GM/CCL composite barriers. As reported in Section 7.2.5, average
measured percolation rates for test plots containing seamed GMs averaged 1.3 mm/year, with the
measured percolation being attributed to thermally-induced moisture movement in the CCL, not
leakage through the composite barrier. The results from Dwyer (2001) for the GM/CCL
composite hydraulic barrier are also quite good, even with the eight holes cut in the GM by the
researchers. The average measured percolation rate for this cover system was 0.13 mm over the
three-year monitoring period. Conversely, the percolation rates for the GM/GCL composite
hydraulic barrier reported by Dwyer (2001) are high and may be due to the holes cut into the GM
or other factors described in Section 7.2.6. More data for this test plot are needed, and further
investigation into the percolation mechanisms is underway.
DRAFT - DO NOT CITE OR QUOTE
7-16
-------
7.3.2 GM Barrier Seam Problem Due to Contamination
Calabria and Peggs (1996) described a cover system project in Pennsylvania where a high rate of
HDPE GM barrier seam failures occurred during construction. The 1.0-mm thick textured
HOPE GM was installed over a MSW landfill between November 1994 and March 1995. The
project specifications required that both the inside and outside tracks of GM fusion seam samples
be destructively tested. Initially only the inside track of fusion seam samples was destructively
tested in shear and peel. After about 50% of the GM installation had been approved, based on
passing destructive test results, and the approved portion of the GM had been covered with a soil
layer, it was determined that the outside track of fusion seam samples had not been tested.
Archived fusion seam samples were subsequently obtained and tested. About 60% (i.e., 25 of
42) of the archived seam samples had inside track peel test failures, primarily due to seam
separation exceeding the minimum specified value of 10%. Most of the failures were associated
with four of nine seaming machines and two of nine operators. Fifty percent (i.e., 6 of 12) of the
extrusion seam samples taken from the section of GM not covered with topsoil also failed.
These failure frequencies for fusion and extrusion seam samples do not include samples
collected and tested to isolate poor quality seams. The installer attributed the high seam sample
failure frequencies to certain volatile constituents (i.e., benzene, toluene, ethylbenzene, and
xylenes (BTEX)) in landfill gas being absorbed by the HDPE GM and inhibiting the fabrication
of good seams. However, after the installer sent a new supervisor to the site, the failure rate for
extrusion seams dropped.
Calabria and Peggs (1996) performed an investigation to determine if the amount of BTEX
absorbed by the HDPE GM impacted seam quality at the site. The investigation included
obtaining archived seam samples for destructive testing and microstructural examination and
analyzing GM from the site for BTEX constituents. They also exposed site-specific GM samples
to BTEX, seamed them, and tested them in peel and shear. Calabria and Peggs found that most
of the archived fusion seam samples showed rippling along the seam tracks and extensive
warping. They attributed the ripples to GM overheating (setting the seaming machine
temperature too high and/or speed too low). They attributed the warp to manual adjustment of
the seaming machine to change its direction. They also noted that the GM at the outer edge of
the seam tracks was notched, creating a location where stresses could be concentrated, which
could potentially lead to stress cracking. Other seams had linear features oriented along the
length of the seam in areas of the seams where the GM was shiny and not heated sufficiently to
melt its surface. Calabria and Peggs attributed these linear features to soil particles being
dragged along the seam by the hot wedge of the seaming machine.
Selected seam samples from the installed GM were collected and analyzed for BTEX
constituents and subjected to peel testing. None of the constituents was detected at a
concentration greater than 1 mg/kg. No relationship was found between constituent
concentration and seam failure rate. Site-specific GM samples exposed to BTEX, seamed, and
then tested in peel were found to have good quality seams. Based on their investigation, Calabria
and Peggs concluded that the high failure rate for GM seam samples was predominantly caused
by soil in the seams (i.e., inadequate cleaning prior to seaming). Other causes of failure were
overheating and, for extrusion seams, inadequate grinding. The BTEX absorbed by the GM had
no apparent impact on seam quality. The following lessons can be learned from this case study:
DRAFT - DO NOT CITE OR QUOTE
7-17
-------
• The absorption of relatively low concentrations of BTEX by HDPE GM appears not to
affect the quality of seams subsequently constructed.
• HDPE GM must be thoroughly cleaned along a seam path before the seam is constructed
since dirt in the seam adversely impacts seam integrity.
• Dual track fusion seaming machines are designed to make high quality seams along two
tracks. Both tracks should be destructively tested since failure of one track is generally
indicative of overall seaming problems, and failure of one track can increases the stress in
the adjacent track.
7.3.3 GM Barrier Seam Problem Due to Moisture
In a cover system application in the southeastern U.S., a 0.9-mm thick CSPE-R GM was
installed using a solvent seaming method. The overlap width was 75 to 100 mm. Seaming was
typically performed in the early morning hours from sunrise until 9:00 or 10:00 a.m. so as to
avoid intense heat during the day. As the project progressed, there were observations of
unbonded blisters within the seam area particularly in the afternoon. The blisters varied in size
(from 10 to 50 mm and either circular or elliptical in shape) and were numerous.
Upon sampling and seam testing, it was determined, primarily from the results of peel tests, that
there were indeed unbonded areas within the seam at the locations of the blisters. Microscopic
examination showed that the solvent did not dissolve the resin in these same areas. The reason
for the afternoon observation of the blisters is that the air in the unbonded areas expanded as the
GM temperature increased.
After considerable evaluation, it was concluded that the high relative humidity and resulting
moisture during the evening and early morning left the GM wet. The installation crew was not
diligent in making the opposing surfaces in the area to be bonded completely dry and the
undulating surface of the scrim-reinforced GM contributed to their resistance to drying the GM
using rags or wipes. After the installation crew began using a portable heater to dry the area to
be bonded, the problem was avoided for the remainder of the project. Repair of the seamed areas
with blisters was performed using a 0.3-m wide cap strip over the entire width of the original
seam.
This case history, along with the previous one in Section 7.3.2, emphasizes that, regardless of
seaming method, field seaming of GMs has two paramount requirements: (i) the area to be
seamed must be clean; and (ii) the area to be seamed must be dry.
7.3.4 Temperature Fluctuations During GM Installation
This project involved closure of an industrial hazardous waste landfill in the southeastern U.S.
during hot, mid-summer conditions. Large temperature fluctuations during cover system
installation presented the installer of a 2-mm thick HDPE GM barrier with several challenges.
Daytime temperature fluctuations of 11 to 17 °C were commonly observed during the
installation. The excessive heat made welding conditions difficult. The expansion and
contraction of the GM also caused problems.
DRAFT - DO NOT CITE OR QUOTE
7-18
-------
The closure design required pipe boot penetrations for gas vents and for cover system
geosynthetics to be tied into the existing liner system along the perimeter of the landfill. The
majority of the GM barrier production welds were of the double-track fusion type. The
perimeter tie-in was performed manually using an extrusion welder. Due to high daytime
temperatures, extrusion welded seams for the tie-in had to be cooled immediately after seaming
to ensure that seam separation did not take place prior to extrudate hardening. Water-cooled
towels were used to accelerate hardening of the weld. Extreme care was required to maintain a
continuous weld. In addition, during hotter periods of the day, compensation wrinkles were
added upslope and parallel to the perimeter tie-in. These compensation wrinkles had a tendency
to creep downslope, accumulating at the tie-in, and in some places requiring repair (Figure 7-6a).
During welding of the tie-in, after a cooling rain, the GM barrier contracted sufficiently to pull
on the gas vent pipe boots at the landfill crest. The stress in the GM was sufficient to distort the
gas vent pipes from a vertical to an inclined position (Figure 7-6b). Repairs were made to the
affected pipe penetrations by the installation of additional compensation wrinkles near the pipe
penetrations.
(b)
Figure 7-6. Effect of Temperature Fluctuations During GM Installation: (a) GM Wrinkles at
Sideslope Toe; and (b) GM Contraction after a Rain.
DRAFT - DO NOT CITE OR QUOTE
7-19
-------
The installation of cover system geosynthetics under variable high-temperature conditions, as in
this case history, requires not only an understanding of GM thermal expansion and contraction
characteristics, but also limitations of welding techniques and other factors. To reduce the
effects of temperature for these types of conditions, the design engineer can specify GMs with
lower coefficients of thermal expansion, light colored GMs, or provisions for keeping dark
colored GMs covered with temporary light-colored protection (e.g., light-colored GTs) at all
times. Also, the design engineer can specify that GM seaming be performed during relatively
cool periods only (i.e., early morning or evening, under cloudy conditions, etc.).
7.3.5 Fate of GM Wrinkles
A 10-ha landfill vertical expansion in the mid-Atlantic U.S. required installation of geosynthetics
over the existing waste mass. The geosynthetics serve not only as a cover system over the
existing waste but also as a part of the liner system for the new expansion area. The cover
system consisted of, from top to bottom: cover soil; GC drainage layer; and 2-mm thick HDPE
GM barrier.
Close coordination was needed between the geosynthetics installer and the earthwork contractor
during placement of the cover soil over the GC drainage layer. When wrinkles were observed
during the placement of soil over the geosynthetics, spotters were used to "walk out" the wrinkles.
Part way through the installation, the CQA engineer determined that an area of cover soil did not
meet specification. When the non-complying soil was removed and the geosynthetics uncovered,
it was found that the GM wrinkles had persisted and several had folded over and crimped (Figure
7-7). The crimping occurred even though the overburden stress was small, due only to the
protective cover soil. These observations are consistent with the laboratory findings on the fate of
wrinkles presented in Koerner et al. (2002).
Repairs were made to the GM that had been creased during the prior placement of cover soil.
Additional restrictions were then imposed on earthwork operations. Placement of protective
cover soil was restricted to early mornings and evening hours (when the GM was cool and
contracted) to minimize wrinkle formation. Full time spotters and personnel were required to be
present at all times during protective cover placement.
Even though spotters had been used during the initial placement of the cover soil, wrinkles in the
GM were found to occur. This case history further highlights the need to control placement of
soils over geosynthetics and to minimize wrinkling in GMs. It is important to keep GM wrinkles
from folding over since this creates strain concentrations, and hence stress concentrations, in the
GM. It is noted that it has been shown analytically that the size of the wrinkles can be reduced by
increasing the shear strength between the GM and the underlying material (Giroud, 1994).
Therefore, for example, the use of textured rather than smooth GM may reduce the risk that large
wrinkles will form.
DRAFT - DO NOT CITE OR QUOTE
7-20
-------
Figure 7-7. Wrinkles Developed in an HOPE GM and Folded Over During Placement of
Cover Soil.
DRAFT - DO NOT CITE OR QUOTE
7-21
-------
7.4 Slope Stability
7.4.1 Overview
Gross et al. (2002) identified cover system slope instability as the most common type of problem
encountered at landfills. Gross et al. collected available information on cover system slope
stability failures, for which they found:
• four landfills at which cover system slope failures occurred during construction;
• eleven landfills at which cover system slope failures occurred after rainfall or thaw; and
• three landfills at which soil cover damage occurred after an earthquake.
Each of these three types of cover system slope stability problems is discussed below. In
addition, the results from the EPA-sponsored GCL test plot slope stability program are also
discussed.
7.4.2 Cover System Slope Failure During Construction
Cover system slope failures during construction have been described by Paulson (1993),
Boschuk (1991), and Gross et al. (2002). The primary causes of failure were identified as: (i)
placing soil over the sideslope geosynthetics from the top of the slope downward, rather from the
toe of the slope upward; (ii) using unconservative presumed values for critical interface shear
strengths; and (iii) using interface shear strength values from laboratory tests performed under
conditions not representative of the actual field conditions.
At a landfill described by Paulson (1993), the design called for geosynthetic reinforcement to be
installed over a nonwoven GT cushion and then covered with soil. The GT cushion was
underlain by a smooth GM barrier. The reinforcement was to be anchored on the top of the
landfill by covering a length of the reinforcement with soil. Slope stability analyses conducted
during design assumed that soil would be placed over the reinforcement from the bottom of the
slopes upward, after the reinforcement had been anchored. However, this requirement was not
incorporated into the construction specifications. When construction began, access to the bottom
of the slope was not available, so the contractor started placing soil from the crest of the slope
downwards. Shortly afterwards, a section of cover system involving the soil, reinforcement, and
GT cushion slid along the interface between the GT and the underlying GM barrier. The main
factor leading to the failure was placement of cover soil from the top down. Moreover, the
construction specifications did not place any limitations on the size or ground pressure of the
construction equipment used, nor on its mode of operation. Consistent with the recommendation
of Daniel and Koerner (1993), soil layers should normally be placed over geosynthetics from the
toe of slope upward to minimize construction-induced tension in the geosynthetics and take
advantage of passive soil resistance at the toe of slope.
At a landfill described by Boschuk (1991), a gravel drainage layer placed on top of a smooth GM
barrier on a 3H: IV slope, slid down the slope, damaging the underlying GM. The contractor had
tried to place the gravel by pushing it up the slope with a bulldozer and then by placing it on the
slope using a clamshell bucket, but neither method worked. Apparently, the drainage layer
material did not develop adequate interface shearing resistance with the underlying GM.
DRAFT - DO NOT CITE OR QUOTE
7-22
-------
Adequate design-phase interface shear testing and slope stability analyses with materials
representative of final construction would have prevented this problem.
At another landfill discussed by Boschuk (1991), as soil was being placed over the already-
installed sand drainage layer on a 3H:1V a slope, the sand slid downslope over a heatbonded
nonwoven GT. Apparently the sand was too coarse to penetrate into the heatbonded GT
openings. Project-specific interface direct shear tests between the sand and GT performed prior
to construction resulted in an interface friction angle of about 21° indicating the slope would be
stable. The tests were performed, however, at normal stresses much larger than the actual field
loading condition. Tilt table interface shear tests performed after the failure and at a lower
normal stress representative of field conditions produced a sand/GT interface friction angle of
about 18°. This latter test result indicates marginal slope instability for this interface on a 3H: IV
(18.3°) slope. The cover system was reconstructed with a needlepunched nonwoven GT that had
a higher interface shear strength with sand than the calendered GT. The lesson from this case
study is that interface direct shear tests should be performed under laboratory test conditions
representative of those expected in the field.
Gross et al. (2002) described a project involving closure of 32-m long, 3H:1V landfill sideslopes.
The design called for geogrid reinforcement to be installed over a smooth HDPE GM barrier and
then covered with overlaying soil layers, with the first such layer being a sand drainage layer.
The construction specifications required the reinforcement to be anchored on the top of the
landfill by extending the reinforcement onto the top deck and covering it with the soil layers
prior to placing soil over the reinforcement on the sideslope. Slope stability analyses were
conducted assuming that the soil layers would be placed over the reinforcement from the bottom
of the slope upward. However, this condition was not incorporated into the construction
specifications. When construction began, existing gas wells on the top deck interfered with
geogrid installation. Where the gas wells interfered with installation, the adjacent geogrid strips
stopped short and did not extend back to their full design anchorage length. Access to the
bottom of the sideslopes was limited at some locations due to wetlands near the slope toe. As a
consequence of these conditions, the contractor placed the sand by pushing it from the crest
downward. This mistake was compounded by the fact that the contractor created a sand
stockpile on the slope near the crest. Shortly after sand placement began, the anchored geogrid
layers ruptured at the slope crest beneath the sand stockpile and construction equipment. The
GM then tore near the slope crest and along outward diagonals down the length of the GM on
both sides of the stockpile. The cover system was subsequently redesigned using textured rather
than smooth HDPE material. The lessons from this case study are that geosynthetics need to be
properly anchored prior to placing soil cover, soils should not be stockpiled on top of
geosynthetics on slopes (unless accounted for in the design), and soil cover should be placed
from the bottom of the slope up.
7.4.3 Cover System Slope Failure After Rainfall or Thaw
Gross et al. (2002) presented case studies of cover system slope failures due to rainfall or
thawing conditions at eleven landfills. The primary causes of failure were identified as: (i) not
accounting for seepage forces; (ii) clogging of the internal drainage layer, which leads to
increased seepage forces; and (iii) not accounting for moisture at the GM/CCL interface (which
DRAFT - DO NOT CITE OR QUOTE
7-23
-------
weakened the interface) due to both rain falling on the CCL surface during construction and
freeze-thaw effects.
Inadequate Design for Seepage Forces: Five cover system slope failures were primarily
attributed to rainfall-induced seepage pressures in soil layers above the failure surface. The
cover systems for the landfills involved in these failures have 3H: IV or 2.5H: IV sideslopes and
are up to about 60 m in slope length. Available details on the cover system failures are given
below.
• Bonaparte et al. (1996) and Vander Linde et al. (2002) described the failure of a cover
system for a landfill in north Georgia. The cover system consisted of a 0.3-m thick
topsoil layer over a stitch-bonded reinforced GCL barrier. Sideslopes were 3H: IV and
up to 54 m in length. The cover system did not have an internal drainage layer and was
designed without consideration of rainfall-induced seepage forces in the topsoil layer.
Construction of the system was completed in the fall of 1994. During the winter of 1995,
the cover system experienced several episodes of downslope movement. Each episode of
movement was immediately preceded by a significant rainfall event. The nature of the
slope movement is illustrated in Figure 7-8 and photographs of the failure are presented
in Figure 6-1. Analyses performed after the failure demonstrated substantial seepage
force buildup due to rainfall, resulting in a calculated factor of safety of less than 1.0 for
sliding of the topsoil layer on top of the GCL. The main lesson from this case study is
that seepage forces should be considered in evaluating cover system stability. When
seepage forces are accounted for, they will typically lead the design engineer to
incorporate an internal drainage layer into the cover system design whenever a
conventional design approach (involving hydraulic barriers and maximum slopes in the
range of 4H: IV to 3H:1V) is used.
• Boschuk (1991) described a project involving a cover system on a 3H: IV slope. The
cover system consisted of, from top of bottom: topsoil layer; medium-coarse sand
drainage layer; woven GT reinforcement layer; and GM barrier. Project-specific
interface shear testing was not performed. The design engineer assumed a sand/GT
interface friction angle of 24°, or about two-thirds of the sand angle of internal friction.
The sand slid on the underlying GT after a rainfall event estimated by Boschuk to have a
two-year recurrence interval. Gross et al. (2002) calculated slope stability factors of
safely of 1.34, 0.98, and 0.63 for this project assuming infinite slope conditions, a 24°
interface friction angle and, respectively, conditions of no seepage force, seepage in the
sand layer, and full seepage in the sand and overlying topsoil layer. The main lesson
from this case study, like the previous one, is that seepage forces should be accounted for
in evaluating cover system stability. A secondary lesson from this case study is that
project-specific interface shear testing should be performed.
• Boschuk (1991) described an additional cover system failure where the primary causes of
failure were inadequate (or no) consideration of seepage forces and/or inadequate
characterization of interface shear strengths. The cover system cross section consisted of,
from top to bottom, topsoil layer, sand drainage layer, and GM barrier. The sand
drainage layer had a specified minimum hydraulic conductivity of 1 x 10"4 m/s. The type
of GM is not identified in the case study. Sliding occurred along the sand/GM interface
after three days of rainfall. A steady-seepage infinite slope analysis was conducted by
DRAFT - DO NOT CITE OR QUOTE
7-24
-------
Gross et al. (2002) for this case study. In their analysis, a secant friction angle of 20° was
assumed for the sand/GM interface. The calculated slope stability factors of safely are
1.09 and 0.80, respectively, without and with full seepage forces in the sand layer. A
lesson from this case study is that sand drainage layers with a hydraulic conductivity of 1
x 10"4 m/s may not be permeable enough to convey flow without the buildup of seepage
forces. A higher permeability drainage medium would perform better.
Soong and Koerner (1997) described the 1995 failure of a cover system on a 40-m long,
2.5H: IV slope that occurred after a heavy rainfall. The cover system consists of a 0.75-m
thick silty sand layer (approximate hydraulic conductivity of 1 x 10"5 m/s) underlain by a
CCL barrier. About two to three years after the cover system was constructed, the sand
slid downslope over the CCL during a storm. The slide was relatively small and
localized. Soong and Koerner attributed the failures to seepage forces that developed in
the sand layer. An infinite slope analysis was conducted by Gross et al. (2002) for this
case study. In their analysis, the friction angle for the sand was assumed to be 30°. The
calculated slope stability factors of safety are 1.44 and 0.66 without and with full seepage
forces in the sand layer, respectively. The lesson from this case study is similar to the
previous one: cover system internal drainage layers may need to have a hydraulic
conductivity much larger than 1 x 10"5 m/s to prevent significant seepage forces. A
higher permeability drainage medium would perform better.
Soong and Koerner (1997) also described the 1996 failure of a cover system on a 50-m
long, 3H: IV slope. The cover system consists of a 0.6-m thick topsoil layer overlying a
0.3-m thick sand drainage layer (approximate hydraulic conductivity of 1 x 10"4 m/s),
which in turn overlies a CCL barrier. About five to six years after the cover system was
constructed, the sand slid downslope over the CCL immediately after a storm. At least
four localized slides occurred. Soong and Koerner attributed the slides to relatively high
seepage forces that developed in the cover system because the drainage layer hydraulic
conductivity was too low. The timing of the slides (5 to 6 years after closure) suggest
that clogging of the sand drainage layer may have occurred to some extent. An infinite
slope analysis was conducted by Gross et al. (2002) for this case study. In their analysis,
the friction angle for the sand was assumed to be 30°. The calculated slope stability
factors of safety are 1.73 and 1.40 without and with full seepage forces in the sand layer,
respectively. With seepage forces in the sand and topsoil layers, the calculated factor of
safety is 0.77. The lessons from this case study are that: (i) the hydraulic conductivity of
cover system internal drainage layers may need to be larger than 1 x 10"4 m/s to prevent
significant seepage forces; and (ii) clogging of an internal drainage layer can reduce its
effectiveness. This latter effect is discussed in more detail below.
DRAFT - DO NOT CITE OR QUOTE
7-25
-------
Exposed
GCL
Displaced Block of Soil
Sliding Interface between
-Topsoil and GCL
Vegetation
Compression Ridges
Waste —^=__~^_—
~=L. —^^~ ~— Stitch-Bonded GCL-
Figure 7-8. Observed Failure Mechanisms for Sliding of Soil Layer Over Stitch-Bonded
GCL.
Clogging of Internal Drainage Layer: Clogging of the cover system internal drainage layer can
impair the ability of the layer to freely drain, resulting in a buildup of hydraulic pressure and
failure of the cover system. This mechanism was identified as the primary factor contributing to
slope stability problems at five landfills. Available details on these cover system slope failures
are given below.
• Boschuk (1991) described a cover system slope failure that appeared to be related to
clogging of the sand drainage layer. The cover system consists of, from top to bottom:
topsoil layer; gap-graded sand drainage layer (minimum hydraulic conductivity of
1 x 10"4 m/s); and smooth GM barrier. The cover system slopes ranged from about 50 to
90 m in length. Within one year of the completion of construction, the entire lower third
of the cover system slid downslope along the sand/GM interface. The sand drainage
layer in the slide zone contained significant fines, presumably washed into the sand from
the topsoil layer and the sand in the upper two-thirds of the slope. Boschuk (1991)
indicated that fines migration had so reduced the hydraulic conductivity of the sand
drainage layer at the bottom of the slope that the layer liquefied under the induced
hydraulic head buildup. Lessons learned from this case study are as follows:
o Gap-graded soils are more prone to migration of finer-sized particles (i.e., internal
instability) than well-graded soils. Particle migration may result in clogging of
the soil. Therefore, if gap-graded soils are used as drainage materials, the
potential for particle migration should be evaluated during design.
o A granular soil drainage layer needs to have a filter to protect against migration of
particles from the overlying topsoil or protective soil layer. This aspect of the
design should be performed using available filter criteria (see Chapter 4 of this
DRAFT - DO NOT CITE OR QUOTE
7-26
-------
document) and/or laboratory testing. GT filter layers can be used in the design if
a sand filter layer is not available or is too costly.
o Cover system slopes should always be evaluated for stability using rigorous
analysis methods that consider the anticipated seepage forces and
interface/internal shear strengths applicable to the cover system.
• Boschuk (1991) described a project where the cover system consists of from top to
bottom: topsoil layer; sand drainage layer; and smooth GM barrier. Perforated collection
pipes wrapped with a nonwoven GT filter were installed in the sand drainage layer. After
a period of time, fines clogged the GT at the pipe perforations, hydraulic head built up the
sand drainage layer, and the cover system slid downslope. Failure occurred at the
sand/GM interface, primarily on the lower third of the slope. After the failure, the pipes
were observed to be dry and the surrounding sand saturated. This failure might have
been prevented if the GT filter wrapped around the pipe had been adequately designed.
A thinner, more open, GT, that allows fine soil particles to pass through but which retains
the sand, would have performed better. Much better, however, would have been to not
wrap the pipe in a GT filter at all, but rather to bed the pipe in drainage gravel and place a
properly designed GT filter around the gravel, or to design the system to allow
unimpeded migration of fines through the pipe perforations. The problems associated
with placement of GT filter layers around pipes (as was done in this case study) have
been clearly described by Bass (1986), Koerner et al. (1993), and Giroud (1996).
• Another failure described by Boschuk (1991) involved a cover system consisting of, from
top to bottom: topsoil layer; nonwoven GT filter; gravel drainage layer; and GM barrier.
Over time, the GT became clogged by the topsoil. As a consequence, infiltrating
rainwater did not drain freely from the topsoil into the underlying gravel. Pore pressures
increased in the topsoil layer, and the topsoil slid downslope over the GT. Failure
occurred primarily on the lower third of the slope. Boschuk (1991) did not indicate if
filter design, interface direct shear testing, or a slope stability analysis were performed as
part of the cover system design. The GT should have been designed to be compatible
with the topsoil using filter criteria calculations and/or laboratory testing. Compatibility
between topsoil and GT filter layers should always be carefully evaluated because the
topsoil may have a low degree of internal stability. Internally unstable soils will typically
be poorly graded, with significant fines and little cohesion.
• Soong and Koerner (1997) described the failure of a cover system on a 45-m long, 3H: IV
slope that occurred in 1996. The cover system consists of, from top to bottom: 0.75-m
thick topsoil layer; 0.3-m thick sand drainage layer; and CCL barrier. The design called
for water in the sand drainage layer to flow to the toe of the slope where it would be
collected in a gravel toe drain and then conveyed through a pipe to a discharge point.
The gravel toe drain was not wrapped with a GT filter. Five to six years after the cover
system was constructed, a number of localized slides of the sand over the CCL occurred.
When the gravel toe drain was exhumed, the gravel was found to be very contaminated
with fines, which presumably migrated into the gravel from the overlying sand and
topsoil. Soong and Koerner attributed the failure to relatively high seepage forces that
developed in the cover system after the gravel toe drain became clogged. Lessons
learned from this case study are similar to those learned from the previous case studies.
DRAFT - DO NOT CITE OR QUOTE
7-27
-------
• Soong and Koerner (1997) described the failure of a cover system on a 45-m long,
2.5H: IV slope between benches that occurred in 1996. The cover system consists of,
from top to bottom: 0.6-m thick topsoil surface/protection layer; 0.2-m thick sand
drainage layer; and CCL barrier. The design called for water in the sand drainage layer
to flow to the toe of the slope where it would be collected in a gravel toe drain and then
conveyed through a pipe to a discharge point. The pipe was wrapped with a GT filter.
As with the previous case study, about five years after the cover system was constructed,
a number of small localized slides of the sand over the CCL occurred. When the gravel
toe drain was exhumed, the GT filter layer was found to be clogged with fines at pipe
perforations. The fines presumably migrated to the GT from the sand and topsoil. Soong
and Koerner attributed the failure to hydraulic head that developed in the cover system
after the GT around the pipe became clogged. As previously discussed, wrapping of
perforated pipes in GTs should be avoided if at all possible due to the relative
inefficiency of placing the filter layer at this location and the potential for clogging (Bass
(1986), Koerner et al. (1993), and Giroud (1996)).
Moisture Changes at GM/CCL Interface: Gross et al. (2002) described a case study involving a
cover system for which construction was not completed until late fall. The project site is located
in northern Ohio. The cover system cross section is illustrated in Figure 7-9. During the first
winter after landfill closure, the cover system was covered with snow and the ambient
temperature was below freezing until the spring.
0.45m
GT Filter
(needlepunched
nonwoven)
0.75 mm PVC GM
3 or 4
Figure 7-9. Cover System Cross Section for Northern Ohio Landfill that Underwent a
Slope Failure after Thaw.
A few days after the first spring thaw, the PVC GM component of the cover system slid over the
CCL component on a portion of 4H: IV slope. An initial investigation after the failure revealed
that water could not exit from the sand drainage layer because the lower end of the drainage
DRAFT - DO NOT CITE OR QUOTE
7-28
-------
layer was blocked by ice and snow. As a result, the cause of the slide was initially assumed to be
the buildup of hydraulic head resulting from the thawing of the blocked drainage path. However,
subsequent slope stability analyses demonstrated that seepage forces above the GM would have
had little effect on the factor of safety with respect to a slide that occurs at an interface located
beneath the GM (see Chapter 6 of this guidance document). With seepage forces identified as
only a minor potential contributor to the slope failure, an additional investigation was conducted
to evaluate the shear strength characteristics of the GM/CCL interface and, in particular, the
effect of temperature fluctuations on interface strength. Interface shear tests simulating the
conditions during the winter (-7 °C) followed by thaw (+0.5 °C) showed that the formation of ice
lenses at the GM/CCL interface at below-freezing temperature increased the water content at the
interface during thaw. This resulted in a marked decrease of the interface shear strength after the
thaw, compared to the interface shear strength before freezing. Slope stability calculations
incorporating the results of the interface shear strength testing program showed that the cover
system would be unstable on a 4H: IV slope if the moisture content of the CCL exceeded 23%.
Systematic measurements of field CCL moisture content showed that this moisture content was
likely exceeded in the area where the slide occurred, while the condition was not met in other
areas. This localized effect (i.e., higher water content) was attributed to heavy rainfall that
preceded the installation of the GM in the area where the slide eventually occurred.
The main lessons from this case study is that freeze-thaw cycles have a significant effect on
interface shear strengths. To avoid potential problems, the interface should be located below the
depth of frost penetration. Also, rainfall onto a CCL immediately prior to GM placement can
lead to lower interface strengths than obtained in interface shear tests performed at "as
compacted" moisture contents.
7.4.4 Soil Cover Damage Due to Earthquakes
Loma Prieta Earthquake: The epicenter of the 17 October 1989 (moment magnitude Mw 6.9)
Loma Prieta earthquake was located approximately 16 km northeast of the City of Santa Cruz.
The focal depth was approximately 18 km, with a fault plane dipping about 10 degrees from the
vertical to the west. The Loma Prieta event produced observational data on the seismic
performance of older, unlined solid waste landfills. Orr and Finch (1990), Johnson et al. (1991),
and Buranek and Prasad (1991) reported on post-earthquake inspections of fifteen landfills.
None of the landfills subjected to strong shaking in the Loma Prieta event were instrumented.
The estimated bedrock peak horizontal ground accelerations (PHGA) at the base of the landfills
in the Loma Prieta event ranged from 0.1 g to 0.5 g. All of the post-earthquake damage
investigators reported only minor or moderate damage (as defined by Matasovic et al. (1995)) to
landfills in this event, with the most common damage being cracking of the cover soil on the
landfill slopes and at transitions between waste and natural ground. Johnson et al. (1991) and
Buranek and Prasad (1991) noted that it was often difficult to distinguish between "normal"
cracks induced by waste settlement and/or decomposition and earthquake-induced cracking.
Repair of this type of cover soil cracking is performed regularly as part of routine landfill
maintenance activities. The earthquake induced cracks in the cover soil were repaired by landfill
maintenance crews immediately following the earthquake without disruption to landfill
operations. Orr and Finch (1990) note that some of the landfill gas recovery systems were
temporarily affected by power loss and that there was above-ground pipe breakage at a number
of the landfills impacted by the Loma Prieta earthquake. However, according to these
DRAFT - DO NOT CITE OR QUOTE
7-29
-------
investigations, all landfill gas recovery systems were repaired and back in operation within 24
hours of the earthquake, and there were no reported post-earthquake changes in quantities of
leachate and extracted landfill gas.
Among the landfills closest to the Loma Prieta earthquake zone of fault rupture, observational
data exist for the Guadalupe, Ben Lomond, Kirby Canyon and Santa Cruz landfills. The
estimated bedrock PHGAs for these landfills are 0.43 g, 0.38 g, 0.34 g and 0.30 g, respectively.
As reported by Johnson et al. (1991), even the highest slopes at these landfills, which include
2H: IV slopes up to 45 m high at the Santa Cruz landfill, 3H: IV slopes up to 45 m high at the
Ben Lomond landfill, and 2H: IV slopes up to 75 m high at the Kirby Canyon landfill, performed
well, with only minor cracking (25 to 75 mm in width) of cover soils observed. Only at the
Guadalupe landfill, as reported by Buranek and Prasad (1991), was minor downslope cover soil
movement observed.
Northridge Earthquake: Augello et al. (1995), Matasovic et al. (1995), and Matasovic and
Kavazanjian (1996) documented damage to soil cover materials at three landfills in the 17
January 1994 Northridge earthquake (moment magnitude Mw 6.7). This earthquake occurred on
a blind thrust fault at a depth of approximately 15 km at the northern end of the San Fernando
Valley within the greater Los Angeles area. Estimated PHGA in bedrock at the landfill sites
ranged from 0.20 g to 0.42 g. Consistent with observations in the Loma Prieta earthquake,
damage in the Northridge event was limited to surficial cracking of cover soils occurring
primarily near locations with contrasting seismic response characteristics (e.g., top of waste
adjacent to canyon slopes). At two of the landfills, the cracking was relatively minor. At one
landfill, a major crack occurred near and parallel to a liner system anchor trench. This crack was
about 200-m long, up to 150-mm wide, with the two sides of the crack vertically offset by up to
100 mm. No waste was exposed. At all three landfills, the damage was dealt with as an
operation issue through post-earthquake inspection and repair (i.e., regrading and revegetating
the cracked soil layers).
The main lesson from these case studies is that surficial cracking of soil cover layers, especially
near locations with contrast in seismic response characteristics (e.g., top of waste by sideslopes),
should be anticipated and dealt with as an operation issue through post-earthquake inspection
and maintenance.
7.4.5 Results of EPA GCL Test Plots
Carson et al. (1998), Daniel et al. (1998), and Daniel (2002) describe the results of an evaluation
of 14 GCL field test plots constructed at a landfill test site in Cincinnati, Ohio. The test plots
were designed and constructed as prototype landfill cover systems. The purpose of the test plots
was to evaluate the internal and interface shear strength characteristics of the commercially-
available GCLs under in-service conditions. Five test plots were constructed on a 3H: IV
(nominal) slope, and nine test plots were built on a 2H: 1V (nominal) slope. Plots on the 2H: IV
slope were nominally 20 m long, while those on the 3H: IV slope were 29 m long. All plots were
two GCL panel widths (9 m) wide and were covered with 0.9 m of silty, clayey sand.
A typical cross section of a test plot constructed on a 3H: IV slope is shown in Figure 7-10. In
general, the test plots were constructed with a double-sided textured GM overlying the GCL,
DRAFT - DO NOT CITE OR QUOTE
7-30
-------
which would be typical of a cover system for a landfill. However, GCLs are also used in cover
systems without GMs. Hence, three plots were constructed with no GM. The plots were drained
internally above the GM using a GC (GT/GN/GT) drainage layer or, for the plots that did not
contain a GM, a sand drainage layer.
Crest
Subsoil
Toe
GC Drainage Layer-
GM
GCL
Figure 7-10. Typical Cross Section for 3H:1V Cover System Test Plot in Cincinnati, Ohio
(modified from Carson et al., 1998).
The rationale for selecting the 2H: IV and 3H: IV slope inclinations was as follows. The 3H: IV
slope was selected to be representative of typical cover systems for landfills in use today. In
order to confirm that GCLs are safe against internal failure on 3H: IV slopes, it must be shown
that they are not only stable, but are stable with an adequate factor of safety. Slope stability
analysis methods are discussed in Chapter 6 of this guidance document. As discussed in Section
6.2.6, a minimum acceptable factor of safety (FSm;n) for static stability analyses of 1.5 will often
be appropriate for permanent cover system applications. The ratio of tan|3 for a 2H: IV slope to
tan|3 for a 3H: IV slope is 1.5. Subject to the assumptions listed above, if a GCL is demonstrated
under a given normal stress to be stable on a 2H:1V slope (i.e., FS > 1.0), the same GCL is
demonstrated to be stable on a 3H: IV slope at the same normal stress with FS > 1.5. Therefore,
the 2H:1V slopes were chosen to demonstrate internal stability of GCLs on 3H:1V slopes with
FS > 1.5. However, it was recognized that constructing 2H: IV slopes was pushing the GCLs to
(and possibly beyond) their limits of stability.
DRAFT - DO NOT CITE OR QUOTE
7-31
-------
Woven or Nonwoven GT
Needlepunched
Fibers-
Sc
>dii
jm
Be
jnt
Dni
te
^
(a)
Nonwoven GT
Woven GT-
Sewn Stitche
-\ oewn ouiuies—\
1 \,
SodiurmBentonite
Mixed with! an Adhesive
(b)
Woven GT
~7
Sodium Bentonite
Mixed with an Adhesive
(c)
GM
Figure 7-11. Schematics of GCLs Used in Cover System Test Plots in Cincinnati, Ohio:
(a) Reinforced, GT-Encased, Needlepunched GCL (e.g., Bentofix and
Bentomat); (b) Reinforced, GT-Encased, Stitch-Bonded GCL (e.g., Claymax);
and (c) Unreinforced, GM-Supported GCL (e.g., Gundseal).
Three types of GCLs, shown schematically in Figure 7-11, were used in the test plot program: (i)
reinforced, GT-encased, needlepunched GCLs (e.g., Bentofix and Bentomat); (ii) reinforced GT-
encased, stitch-bonded GCL (e.g., Claymax); and (iii) unreinforced, GM-supported GCL (e.g.,
Gundseal). For the ten test plots in which a GM was placed over the GCL, the GM was a 1.5-
mm thick textured HOPE GM.
Construction of the test plots began on November 15, 1994 and was completed on November 23,
1994. However, one plot (P) was constructed on June 15, 1995. The test plots were first graded
to provide a smooth subgrade. Next geosynthetics were installed by pulling them down from the
crest of the slope (Figure 7-12), and then cover soil was placed (Figure 7-13) by starting at the
bottom of the slope and working upslope. In plots incorporating a GC drainage layer, the GM
and GC were extended beyond the GCL at the toe of the slope and another 1.5m past the end of
DRAFT - DO NOT CITE OR QUOTE
7-32
-------
Figure 7-12. GCL Panels Deployed on Slopes of Cincinnati, Ohio Test Plots by Pulling
Them Downslope from a Spreader Bar at the Slope Crest.
Figure 7-13. Cover Soil on the Cincinnati, Ohio Test Plots Placed over the Geosynthetics
from the Slope Toe Upward.
DRAFT - DO NOT CITE OR QUOTE
7-3
-------
the cover soil (Figure 7-10). For plots constructed with a sand drainage layer, a piece of GC
material was embedded in the sand at the toe of the slope and then extended 1.5 m beyond the
end of the cover soil.
All of the geosynthetic materials in each test plot were brought into their respective anchor
trenches, which were then backfilled. The toe of each test plot was excavated at the completion
of construction so that no buttressing (i.e., passive) force could be mobilized at the toe of the
slope. To prevent the development of tension in the geosynthetic components above the mid-
plane of the GCLs, all components above the mid-plane, including the upper GT of the GCL,
were cut at the crest of the slope (Figure 7-14). Cutting occurred in the spring of 1995, after the
winter thaw and about five months after construction of the test plots. However, the
geosynthetics were not cut in plot P, which was constructed later in the program for the sole
purpose of evaluating hydration of bentonite encased between two GMs.
All geosynthetics above
mid-plane of GCL were cut,
including upper GT or
GM component of GCL
(if present)
-GCL
— GM
1—GC Drainage Layer
Figure 7-14. Cut in Anchor Trench Geosynthetics above Mid-Plane of GCL on Cincinnati,
Ohio Test Plots.
Instrumentation for the test plots included gypsum blocks and fiberglass moisture sensors and
wire displacement gauges (extensiometers). A discussion of moisture sensors is provided in
Chapter 8 of this guidance document.
As described by Carson et al. (1998), Daniel et al. (1998), and Daniel (2002) the test plots were
observed for over three years. A summary of information on the test plots is given in Table 7-6.
A summary of results of the test plot program is given in Table 7-7.
DRAFT - DO NOT CITE OR QUOTE
7-34
-------
Table 7-6. Information on GCL test plots (from Daniel et al., 1998).
Nominal Target Actual Actual Actual
Test Type of Slope Slope Slope Slope Plot
Plot GCL Angle Angle Length Width
(H:V)
A Gundseal 3:1
n
n
(m) (m)
Cross
Section
(Top to
Bottom)1
GCL Side
Facing
Upward
18.4 16.9 28.9
10.5 Soil/GC/GM/GCL Bentonite
GCL Side
Facing
Downward
GM
B
C
D
E
F
G
H
I
J
K
L
M
N
P
Bentomat
ST
Claymax
500SP
Bentofix
NS
Gundseal
Gundseal
Bentomat
ST
Claymax
500SP
Bentofix
NW
Bentomat
ST
Claymax
500SP
Bentofix
NW
Erosion
Control
Bentofix
NS
Gundseal
3:1
3:1
3:1
3:1
2:1
2:1
2:1
2:1
2:1
2:1
2:1
2:1
2:1
2:1
18
18
18
18
26
26
26
26
26
26
26
26
26
26
.4
.4
.4
.4
.6
.6
.6
.6
.6
.6
.6
.6
.6
.6
17.8
17.6
17.5
17.7
23.6
23.5
24.7
24.8
24.8
25.5
24.9
23.5
22.9
24.7
28.
28.
28.
28.
20.
20.
20.
20.
20.
20.
20.
20.
20.
20.
.9
.9
.9
.9
.5
.5
.5
.5
.5
.5
.5
.5
.5
.5
9.0
8.1
9.1
10.5
10.5
9.0
8.1
9.1
9.0
8.1
9.1
7.6
9.1
9.0
Soil/GC/GM/GCL
Soil/GC/GM/GCL
Soil/GC/GM/GCL
Soil/GC/GCL
Soil/GC/GM/GCL
Soil/GC/GM/GCL
Soil/GC/GM/GCL
Soil/GC/GM/GCL
Soil/GT/Sand/GCL
Soil/GT/Sand/GCL
Soil/GT/Sand/GCL
Soil
Soil/GC/GM/GCL
Soil/GC/GM/GCL
Woven GT
Woven GT
Nonwoven
GT
GM
Bentonite
Woven GT
Woven GT
Nonwoven
GT
Woven GT
Woven GT
Nonwoven
GT
No GCL
Nonwoven
GT
Bentonite
Nonwoven
GT
Woven GT
Woven GT
Bentonite
GM
Nonwoven
GT
Woven GT
Nonwoven
GT
Nonwoven
GT
Woven GT
Nonwoven
GT
No GCL
Woven GT
GM
1GC = GT/GN/GT.
All test plots were initially stable, but over time as the bentonite in the GCLs became hydrated,
three slides (all on 2H: IV slopes) involving GCLs occurred. One slide involved an unreinforced
GCL in which bentonite that was encased between two GMs unexpectedly became hydrated.
The other two slides occurred on 2H: IV slopes at the interface between the woven GT
components of the GCLs and the overlying textured HDPE GMs. A photograph of the test plots
at which these two interface slides occurred is presented in Figure 7-15.
DRAFT - DO NOT CITE OR QUOTE
7-35
-------
Table 7-7. Summary of calculated factor of safety (FS) and actual slope stability (from
Daniel et al., 1998).
Large-
Test Plot Slope Peak Displacement Peak
Designation Angle Friction Angle Friction Angle FS
Large- GCL
Displacement Performance
FS
A
B
C
D
E
F
G
H
I
J
K
L
N
P
16.9
17.8
17.6
17.5
17.7
23.6
23.5
24.7
24.8
24.8
25.5
24.9
22.9
24.7
yjw
231
201
291
202(H)
202(H)
231
201
371
-31 1
313
-31 1
-371
202(H)
352(D)
211
201
221
202(H)
202(H)
211
201
241
-311
313
-311
2_52(D)
1.31
1.11
1.81
2(D)
2.3
1.21
1.31
1.1
2(H)
-24
20
,2(H)
0.82(H)
1.01
0.81
1.61
1.31
1.31
1.31
1.81
0.82(H)
1.1
0.8
2(H)
2(H)
0.9'
0.81
1.01
1.31
1.31
1.31
1.1
0.8
2(H)
Stable
Stable
Stable
Stable
Stable
Internal Slide
Interface Slide
Interface Slide
Stable4
Stable4
Stable4
Stable4
Stable
Stable
1 GM/GCL interface
2 Internal GCL strength for dry (D) or hydrated (H) bentonite
3GCL/drainage sand interface
4 Large displacement occurred in subsoil below GCL, but not in or at the interface with GCL
As discussed by Daniel et al. (1998), the experience from these test plots provides several
conclusions of practical significance to engineers. At the low normal stresses associated with
cover systems, the interface shear strength is generally lower than the internal shear strength of
internally-reinforced GCLs. The weakest interface will typically be between a woven GT
component of a GCL and the adjacent material, which in this case was a textured HDPE GM.
The interface strength may be low in part because of the tendency of bentonite to extrude
through the openings in the relatively thin, woven GT and then into the interface as the GCL
hydrates. Design engineers are encouraged to consider GCLs with relatively thick, nonwoven
GT components in critical situations where high interface shear strength is required.
Current engineering practice for evaluating the stability of GCLs on slopes is to conduct direct
shear tests and then to use LE methods of slope stability analysis to calculate factors of safety
using the results of those tests. This approach was described in detail in Chapter 6 of this
document. The experience from the test plot program has validated this approach. All three test
plots that slid had calculated factors of safety less than 1.0. All remaining (stable) test plots had
factors of safety greater than 1.0. Based on the experience from this study, cover systems
containing GCLs cannot achieve slope stability factors of safety normally considered adequate
on 2H: IV slopes. It appears, however, that 3H: 1V slopes (depending on materials) can be
constructed with factors of safety of at least 1.5 for the conditions existing in this project.
DRAFT - DO NOT CITE OR QUOTE
7-36
-------
Figure 7-15. Plots G (Left) and H (Right) Approximately Two Months After Construction
and Several Days after the Slide in Plot G.
7.5 Waste Settlement
Gross et al. (2002) described a project in which landfill settlement caused tearing of cover
system GM boots around gas well penetrations. The landfill cover system has a 1-mm thick
HDPE GM barrier and was constructed in 1991 and 1992. By late 1992, a gas collection system,
including vertical HDPE gas collection wells that penetrate the GM barrier, had been installed in
the landfill. At each penetration, an HDPE GM boot was clamped to the well and extrusion
welded to the GM barrier to seal the barrier around the well. When several of the GM boots
around the wells were inspected in 1995, the boots were observed to be torn from the GM
barrier. The boots were not designed to accommodate settlement of the waste, which would
cause downward displacement of the GM barrier relative to the wells. Since the cover system
had been constructed, the landfill top deck had settled from 0.3 to 0.9 m. The problem was
resolved by replacing the gas extraction well boots with new expandable boots that can elongate
up to 0.3 m. These boots can also be periodically moved down the well to accommodate landfill
settlement. The lesson from this case study is that GM boots in cover systems must be designed
to accommodate landfill settlements.
Another example of the impacts of settlements on a cover system is shown in the photographs in
Figure 7-16. Surface tension cracks caused by differential settlement of underlying MSW are
clearly evident. The cracks occurred in a soil cover system at an arid site in the western U.S.
DRAFT - DO NOT CITE OR QUOTE
7-37
-------
;,.,-,
Figure 7-16. Surface Tension Cracks in Cover Soils from Differential Settlement of
Underlying MSW.
DRAFT - DO NOT CITE OR QUOTE
7-38
-------
The soil material used to construct the cover consists of a silty, gravelly sand, intended to be
resistant to erosion and desiccation cracking. These cracks were observed to occur throughout
the cover system, with most aligned perpendicular to the slope (constant elevation). The cracks
were observed to act as drains for surface runoff during infrequent storm events, allowing
percolation into the waste mass.
7.6 Stormwater Management and Erosion Control
7.6.1 Failure of Erosion-Mat Lined Downchute
Harris et al. (1992) described the failure of a geosynthetic erosion mat-lined downchute on the
cover system of a landfill in Missouri. An erosion mat was used to line one downchute that
conveyed runoff from approximately 2 ha of cover system and 8 ha of adjacent property; riprap
was used to line the remaining three downchutes that drained a total of about 10 ha. The erosion
mat consisted of a polyethylene, three-dimensional, turf reinforcement mat (TRM). The mat-
lined downchute was installed on the top deck, starting in about 3 m from the slope crest, down
the sideslope, and along a perimeter section of the landfill. At the inlet, the downchute slope is
about 5%, and runoff is diverted into the downchute by small diversion berms. The downchute
grade increases to 33% on the sideslope. Near the slope toe, the downchute has a more gentle
inclination of about 8%. Riprap was placed in the downchute at this lower slope transition for
energy dissipation. TRM was supplied in rolls that were 1.5 m wide and 30 m long. Adjacent
rolls were overlapped at least 75 mm and secured to the underlying soil with 200-mm long
staples installed at 0.75 m spacings. Roll ends overlapped a minimum of 0.45 m and were
shingled downward. TRM was also anchored in 0.3-m deep trenches at the top of each roll and
along the sides of the downchute. After the mat was placed, grass seed was applied and covered
with about 13 mm of topsoil. Within one month after construction, following a series of
significant rainfall events, the channel was unserviceable. Soil had raveled along the sides of the
downchute, soil had eroded underneath the mat and along mat panel overlaps, and the mat had
moved downslope about 2 m. Failure of the mat appeared to have started at the top of the slope
and progressed downward. Though grass was becoming established across the cover system by
this time, there was little grass in the downchute at the time of failure.
The most severe damage to the downchute is believed to have occurred after a peak rainfall
intensity of about 64 mm/hr, estimated to represent a 1-hr storm with a 5-year recurrence
interval. The peak runoff from this storm in the downchute on the sideslope was estimated by
Harris et al. (1992) to be 1.33 m3/s. The corresponding peak velocity in the downchute was
calculated to be 2.9 m/s. After the failure, a detailed laboratory testing program was conducted
to evaluate the relationship between flow velocity and erosion of a mat-lined surface for a
simulated flow duration of 0.5 hr. The results of the study indicated that fully-grassed, mat-lined
channels had noticeable erosion at flow velocities of about 5 m/s. However, without grass, the
velocity required to develop noticeable erosion was about 3 m/s. Harris et al. (1992) concluded
that the combination of large drainage area, steep slope, and the inability of grass to sprout
quickly in the channel lead to failure of the downchute.
DRAFT - DO NOT CITE OR QUOTE
7-39
-------
Based on the information in Harris et al. (1992), the following lessons can be learned from this
case study:
• Flow velocities in drainage channels under the design storm should be calculated so the
appropriate channel lining can be selected. If an erosion mat is selected for a channel and
the erosion mat cannot withstand the design flow velocities until grass is established,
significant maintenance and/or failure of the downchute should be anticipated.
• If the downchute had been constructed earlier, within the plant growing season, the grass
may have become established faster and erosion of the downchute may have been less
severe. The mat was installed and seeded in the fall, when plant growth is relatively
slow, resulting in an extended period with poor to no grass cover in the downchute. The
average plant growing season at the site starts in April and ends in October, the month in
which construction of the downchute was completed. Every effort should be made to
establish cover system vegetation prior to the onset of cool fall weather.
7.6.2 Excessive Erosion and Gullying
Gross et al. (2002) described a 16 ha landfill cover system, with 60-m long slopes inclined at
3H:1V. The design called for sand berms to divert surface-water runoff from the top deck of the
landfill to six riprap-lined downchutes on the landfill sideslopes. Sand diversion berms were
also located at a few locations on the sideslopes. The cover system consists of the following
components, from top to bottom:
• vegetated topsoil layer, 0.2 m thick on the top deck and 0.3 m thick on the sideslopes;
• sand drainage layer with a specified minimum hydraulic conductivity of 1 x 10"5 m/s, 0.2
m thick on the top deck and 0.4 m thick on the sideslopes; and
• 1 -mm thick HOPE GM barrier.
Within three years after construction, about 0.8 ha of the cover system was severely eroded and
about 0.1 ha of cover soil had slid downslope. Sixteen deep gullies developed on the landfill
sideslopes in the vicinity of the riprap-lined downchutes and in areas where the sand berms at the
slope crest had been breached due to differential settlement and sheet flow concentration on the
top deck. Gullies typically started near the slope crest and propagated downslope. The gullies
extended through the topsoil and sand drainage layers down to the GM barrier (Figure 7-17). In
two areas, major sliding of the topsoil and sand drainage layers occurred. In several locations,
the GM was damaged by punctures and tears, and the subgrade beneath the GM was irregular.
EPA HELP model simulations conducted after the erosion was observed indicated that the sand
drainage layer had insufficient capacity to convey surface-water infiltration from the 25-year, 24-
hour storm. Under this condition, the flow that could not be conveyed within the drainage layer
backed-up into the overlying topsoil layer and as surface flow. Seepage forces in the sand
drainage layer and topsoil layer reduced slope stability and increased surface erosion. Other
project details that contributed to the development of erosion and gullies at the site include: (i)
sand diversion berms and downchutes were designed such that they did not intercept lateral flow
in the sand drainage layer; (ii) runoff collected by berms and downchutes could infiltrate through
the topsoil layer and enter the drainage layer; and (iii) a lack of access control resulted in
unauthorized trafficking of four-wheel drive vehicles and dirt bikes on the landfill. The
following lessons can be learned from this case study:
DRAFT - DO NOT CITE OR QUOTE
7-40
-------
• The surface-water runoff management strategy for this landfill, which did not result in
diversion of internal drainage from the top deck to the downchutes and allowed
uninterrupted sheet flow over the 60-m long, 3H:1V sideslopes, proved inadequate to
prevent surface erosion and localized slope instability. A design that incorporated both
drainage layer interceptors and surface-water runoff interceptors (such as benches or
swales) on the sideslopes would likely have been more effective in limiting erosion and
localized failure.
• Design analyses for this facility did not adequately characterize potential peak flows in
the sand drainage layer. For future projects, it is recommended that the guidance given in
this document be used to estimate the required flow capacity. Also, as previously
discussed in this document, a hydraulic conductivity of 1 x 10"5 m/s for a cover system
drainage layer is too low for many applications, including this case study. Hydraulic
conductivity values in the range of 1 x 10"3 m/s, or even higher, will often be necessary to
allow unimpeded drainage while minimizing the build-up of seepage forces in the
sideslope.
• Design of the drainage layer at slope transitions (e.g., drain outlets and benches) is
critical to the effective functioning of the drainage layer. If not properly designed, flow
will back up and generate hydraulic pressure at the slope transition. For flow not to back
up in a drainage layer flowing full, flow capacity across the slope transition must not
decrease. Chapter 4 of this document provides guidance on the design of internal
drainage layers at slope transitions and outlets.
Figure 7-17. Deep Gullies Through the Topsoil and Sand Drainage Layers Exposed the
GM Barrier on 60-m Long, 31-1:1 V Landfill Sideslopes.
DRAFT - DO NOT CITE OR QUOTE
7-41
-------
Figure 7-18 presents a photograph (Dwyer 1997) of a RCRA Subtitle D cover system for a
closed MSW landfill in an arid area of the western U.S., with no surface-water runoff control
system for the portion of the landfill shown in the photograph. Erosion gullies can be clearly
seen in the photograph. These gullies were formed by a single intense storm. The gullies were
deep enough to cut through the entire cover, exposing waste. The cover system sideslopes are
about 3H: IV and the surface layer consists of a silty, gravelly sand.
*_r* - -
Figure 7-18. Gullies on a RCRA Subtitle D Cover System without Surface-Water Runoff
Control System and Located in an Arid Setting.
7.6.3 Failure of Surface-Water Runoff Collector
Figure 7-19 presents photographs of a failed surface-water runoff control system for another
portion of the closed MSW landfill at an arid climate site described above. The control system
consists of corrugated metal pipe-arch culverts installed in the cover system to both intercept
downslope surface-water runoff (culverts placed perpendicular to slope) and convey collected
runoff to a designated collection area (culverts placed in downslope direction). The design basis
for the culverts is not known. The hydraulic capacity of the culverts was not adequate to contain
the runoff and overflow occurred during previous storm events. It appears that the sides of the
culvert blocked entry of runoff into the culvert, causing surface water to flow parallel to the
culvert. This resulted in erosion adjacent to, and beneath, the culvert, exposing waste. Also, the
culverts were prone to siltation and infilling. The lessons learned from this case study are that
non-vegetated cover soils must be designed to convey surface-water runoff without excessive
erosion, runoff interceptors and conveyance structures must be adequately sized, and inlets to the
structures must be designed to not impede the inflow and cause erosion around the structure.
DRAFT - DO NOT CITE OR QUOTE
7-42
-------
•
Figure 7-19. Failed Surface-Water Runoff Control System for Another Portion of the
Closed MSW Landfill Located at an Arid Climate Site and Shown in Figure
7-18.
DRAFT - DO NOT CITE OR QUOTE
7-43
-------
7.7 Gas Pressures
A 40 ha closed MSW landfill was being utilized for gas recovery via deep wells that were placed
at approximately 100 m spacings. The cover system included a GM barrier overlain by a
combined 1 m thickness of various soil layers (Figure 7-20). The well perforations began 8 to 10
m below the cover system (i.e., the upper portion of the wells were not perforated). As a
consequence of the wide well spacing combined with the absence of perforations in the upper
part of the well, gas generated in the upper portion of the landfill accumulated beneath the cover
system, generating uplift pressures on the underside of the GM. As the gas pressure beneath the
GM increased, the normal stress, and, thus, the shear strength, between the underside of the GM
and the GT beneath it decreased (Figure 7-20). This resulted in the GM and overlying materials
gradually moving downslope. The GM and overlying GC (GT/GN/GT) strained considerably at
the top of the slope (Figure 7-21) and folded over at the toe of the slope (Figure 7-22). Tension
cracks were also evident at the top of the slope, and bulging of vegetation, cover soil, and
geosynthetics was apparent at the toe of the slope. The lesson from this case study is that the
spacing of gas extraction wells must be close enough to prevent the buildup of gas pressure on
the underside of the cover system. Also, well perforations must not be so deep as to create a
"dead zone" with respect to gas collection beneath the cover system. In some cases, a granular
or geosynthetic gas transmission layer should be used to provide for more efficient movement of
landfill gas to well locations.
Gas Extraction
See Figure 7-21- x
Vegetatio
Extraction Well
(nonperforated section)
(8-10m)
Extraction Well
(perforated section)
Figure 7-20. Gas Pressures Built Up Beneath Cover System of Closed MSW Landfill
Because Upper Portion of Gas Extraction Wells Was Not Perforated.
DRAFT - DO NOT CITE OR QUOTE
7-44
-------
Tension Crack
GC (GT/GN/GT)
Drainage Layer
GM
GT
Slippage at GM/GT Interface
Figure 7-21. Gas Pressures Beneath Cover System GM Resulted in Slippage at GM/GT
Interface with Straining of the GM and GC at the Slope Crest.
Cover Soil with
Bulge at Toe
GC Drainage Layer
Crimped and Wrinkled GM
Figure 7-22. Slippage at GM/GT Interface Also Caused the GM and GC to Fold Over at the
Slope Toe.
DRAFT - DO NOT CITE OR QUOTE
7-45
-------
At the extreme, gas uplift pressure can be so great as to cause the GM to push the cover soil
aside and expand into a large "whale" as shown in Figure 7-23. The reader should note that this
photograph was not taken at the site of the case study described above, but (in other cases) this
extreme situation has been observed to occur. Landfill gas, if not collected, will also impact
cover systems that do not contain GMs. Vegetation on many landfill cover systems has been
killed by landfill gas emissions. Figure 7-24 presents a photograph (Dwyer 1997) illustrating
this problem. Even in arid climates with non-vegetated surface layers, the impacts of gas
migration can be evident. Figure 7-25 shows surface stains produced by landfill gas throughout
the cover of a closed MSW landfill.
Figure 7-23. GM "Whale" Caused by Gas Pressures Beneath the GM.
7.8 Miscellaneous Problems
Gross et al. (2002) described a project in New York involving the inadvertent use of
contaminated topsoil. During placement of the topsoil layer for a landfill cover system, several
truckloads of soil brought to the site by the contractor had an aromatic odor. The project
specification for topsoil prohibited deleterious material in the topsoil, so topsoil hauling was
ceased until the affected soil could be tested. Samples of the affected soil were collected and
analyzed for VOCs and metals. Based on the results of the testing, the soil was found to contain
unacceptably high concentrations of lead. Topsoil that smelled aromatic or contained chemicals
ionized by a photoionization detector was removed from the site. Each truckload of topsoil
subsequently brought to the site was screened using the above criteria. EPA recommends that
soil borrow sources be investigated by the owner unless the materials are supplied by a
commercial materials company (Daniel and Koerner, 1993). In the case study described above,
topsoil was excavated by the contractor from an off-site property. If the owner had required that
test pits be excavated so the topsoil could be inspected prior to construction, the topsoil
contamination may have been identified earlier. The soil contamination also might have been
identified earlier if the contractor had been required to submit chemical analyses on samples of
the topsoil brought to the site.
DRAFT - DO NOT CITE OR QUOTE
7-46
-------
Figure 7-24. Landfill Cover System Vegetation Killed by Landfill Gas.
DRAFT - DO NOT CITE OR QUOTE
7-47
-------
Figure 7-25. Surface Stains on Landfill Cover System Caused by Landfill Gas.
DRAFT - DO NOT CITE OR QUOTE
7-48
-------
Chapter 8
Performance Monitoring
8.1 Introduction
Performance monitoring of cover systems is necessary to both satisfy regulatory requirements
and confirm the performance of a cover system. The feedback on the effectiveness of a cover
system design can also improve future designs and performance predictions. As discussed by
Kavazanjian (2000), development of performance monitoring data for geoenvironmental projects
(such as cover systems) is often complicated by a number of factors:
long time periods of interest;
. imperfect knowledge of phenomena and impacts, which is sometimes addressed by multi-
parameter modeling, periodic review and updating of monitoring plans, and sensitivity
studies;
. measurement of very small quantities or changes in a physical system, a factor that has
resulted in the development of improved monitoring techniques and methods of statistical
analysis and in the monitoring of surrogates; and
. difficulty in measuring parameters of interest, which is sometimes addressed by making
indirect measurements (e.g., monitoring soil moisture rather than percolation through the
cover system) or monitoring surrogates.
For MSW landfills and HW facilities, post-closure monitoring is required to assure that post-
closure care needs are identified and addressed. Regulations for MSW landfills presented in 40
CFR 258.61(c) and regulations for FIW facilities presented in 40 CFR 264.118 require facility
owners or operators to prepare a written post-closure plan that includes a description of the
performance monitoring activities and the frequency of such activities. The post-closure care
period of 30 years given in RCRA regulations has generally been considered by EPA to be the
minimum timeframe for performance monitoring and maintenance. EPA has the authority to
designate a longer post-closure period under 258.61(b) if necessary for the continuing protection
of human health and the environment. Requirements analogous to those given above for MSW
landfills exist for FIW disposal facilities in 40 CFR 264.117(a)(2)(ii). Also, the post-closure
monitoring requirements for MSW and/or FIW landfills will likely also be ARARs for any cover
system that forms part of a CERCLA remediation. In addition, Five-Year Reviews may need to
be performed, as described in "Comprehensive Five-Year Review Guidance" (June 2001)
OSWER 9355.7-03B-P, EPA 540-R-01-007.
While performance monitoring is important for all facilities with a cover system, it is particularly
such for closed facilities, such as old dumps and remediation sites, not underlain by engineered
liner systems or leachate collection systems which themselves can be monitored. For these sites,
percolation monitoring via a lysimeter (see Section 8.2.4) or soil moisture monitoring (see
DRAFT - DO NOT CITE OR QUOTE
8-1
-------
Section 8.3) is recommended. Such monitoring is also recommended for alternative cover
systems, including those with ET or capillary barriers (see Sections 3.2 and 3.3, respectively).
Prior to implementing a monitoring program, it is important to establish the criteria (i.e., action
levels) for acceptable performance. These criteria are typically developed on a project-specific
basis and may consider the characteristics of the material being contained, human health and
environmental risk, properties of the cover system components, hydrogeologic setting, and other
factors. For example, as discussed in Section 1.2.3, EPA requires that a landfill cover system
have a maximum percolation rate over the considered monitoring period to prevent the "bathtub"
effect. Exceedance of site specific percolation criteria could trigger additional requirements for
the landfill owner or operator. For example, the facility owner could perform an investigation of
the higher than anticipated percolation rates, with the study including an assessment of the
monitoring instrument accuracy and drift, condition of the in-place cover system components,
anticipated performance based on modeling (maybe there was a significant weather event), and
other tasks.
In a project-specific context, monitoring will provide the facility owner/operator, design
engineer, regulators, and other stakeholders with the data necessary to evaluate whether project
design criteria are being achieved. For the entire industry, additional data on the hydraulic and
geotechnical performance of cover systems would be very beneficial to the development of
improved materials, designs, construction procedures, and monitoring/maintenance procedures
for these types of facilities. As previously noted in this guidance document, few data currently
exist on the field hydraulic performance of cover systems and on their long-term structural
integrity when subjected to total and differential settlements.
The types of monitoring systems addressed in this chapter are:
• infiltration monitoring systems (Section 8.2);
• soil moisture monitoring systems (Section 8.3);
• gas emissions monitoring systems (Section 8.4); and
• settlement monitoring systems (Section 8.5).
Other types of post-closure monitoring activities typically associated with waste containment
facilities are not addressed herein. These include groundwater monitoring systems, landfill gas
monitoring systems, and monitoring for physical conditions at the site, such as the condition of
vegetative cover, erosion control structures, sediment control structures, leachate collection and
removal system, landfill gas extraction system, etc. The condition of all of these latter systems
and structures must be monitored during the post-closure period to assure adequate performance
of the site in the long term and to comply with various regulatory requirements. These systems
all require regular inspection and maintenance, topics which are addressed in Chapter 9.
DRAFT - DO NOT CITE OR QUOTE
8-2
-------
8.2 Infiltration Monitoring
8.2.1 Overview
Infiltration monitoring can be performed indirectly, by monitoring leachate collection system
flows for landfills containing such systems, or more directly, by monitoring the cover system
internal drainage layer when one exists or by monitoring a lysimeter installed beneath the
hydraulic barrier layer. Each of these techniques is described below.
8.2.2 Leachate Collection System Monitoring
Data on the quantity and composition of leachate generated within a landfill can provide
significant insight into the performance of a cover system. In facilities underlain by a leachate
collection system and composite liner, leachate flow data can be used as an indicator of cover
system performance.
If a cover system is properly designed and installed, the rate of leachate flow into the leachate
collection system will decrease with time, with the possible exception of cases where leachate
recirculation is practiced. This trend is clearly seen in Figures 1-8 and 1-9. For a cover system
designed and installed to prevent infiltration, the long-term leachate collection system flow rate
would be expected to approach zero. If the cover system does not act as an effective hydraulic
barrier, higher than anticipated long-term leachate flow rates might be observed.
The decrease in leachate collection system flow rate with time after closure can occur relatively
rapidly (e.g., within a few months) in some cases or more slowly (e.g., over several years),
depending on the type and thickness of waste, the waste's moisture content relative to its field
capacity at the time of closure, and, to a lesser extent, the rate of waste degradation (for MSW).
Evaluation of cover system performance during this transition period requires some judgment.
Techniques that can be used to help in the evaluation include: (i) plotting leachate flow rates in
the manner shown in Figures 1-8 and 1-9 to observe the time trend in flow rates; (ii) estimating
the timeframe for residual drainage from the waste using Darcy's equation, the known thickness
of the waste, an estimate of the unsaturated hydraulic conductivities of the waste and
daily/interim cover materials, and an assumed hydraulic gradient equal to one; and (iii) looking
for anomalies in the trend of leachate flow rate with time.
With respect to item (ii) above, timeframes for residual drainage calculated using Darcy's
equation should be considered at best order-of magnitude estimates because of the difficulty in
estimating an appropriate value for the unsaturated hydraulic conductivity of waste, the fact that
the unsaturated hydraulic conductivity of waste is not constant but rather varies with matric
potential, and the difficulty in accounting for such factors as channelized flow along preferential
pathways in the waste and lateral flow at interfaces between waste and daily/interim cover
layers. Based on experience, it appears that the use of Darcy's equation, coupled with published
estimates for MSW waste permeability, will typically provide a conservative, overestimate for
DRAFT - DO NOT CITE OR QUOTE
-------
the timeframe for residual drainage of MSW. For example, for a 25-m thick MSW landfill with
an assumed average unsaturated hydraulic conductivity of 1 x 10"7 m/s, the timeframe for
residual drainage from the waste by gravity is 7.9 years.
With respect to item (iii) above, several different types of anomalies in flow rates can occur. If
periodic increased leachate collection system flow rates are observed, the timing of the increases
should be compared to the timing of precipitation events at the project site. A correlation
between the two is potentially indicative of a breach in the cover system. The most common
potential breach locations are around gas well penetrations through the cover system and at the
edge of the cover system around the perimeter of the facility. If the increased flow rates were
due to a breach and associated influx of precipitation, the concentrations of leachate constituents
in the leachate collection system flow would also be expected to lower (i.e., the flow is more
dilute) during the period of increased flow than during other periods. If the increased flow rates
do not correlate with precipitation, other sources need to be investigated. Another potential
source involves the release of a slug of leachate from the waste to the leachate collection system.
If the source of the flow anomalies is slug flow along preferential pathways in the waste (as
opposed to uniform, porous media-type flow), leachate constituents would be expected to be
similar to those at earlier times. However, the constituent concentrations may potentially be
lower during the anomaly than at earlier times if a significant amount of the teachable
constituents have already been transported from the pathways. Another potential source of long-
term leachate flow for some older landfills is groundwater infiltration, either from perched water
zones or from a continuous zone of saturation that rises above the bottom of the facility.
Indicators of groundwater infiltration include relatively dilute leachate chemistry, changes in
leachate predominant ion chemistry, and correlation between leachate collection system flow
rates and changes in groundwater levels at the site.
It should be noted that the observation of a reduction in leachate collection system flow rate with
time after closure does not by itself prove that a cover system is functioning as designed. The
observed reduction in flow rate after closure may be due to decreasing residual drainage from the
waste, with percolation into the waste reduced from the pre-closure value, but still at a rate above
the intended design value. A slow rate of percolation into the waste may not be reflected in
leachate collection system flow rates for some period of time, due to available moisture
absorption capacity of some or all of the solid waste mass.
In summary, monitoring of flow from the leachate collection system is extremely valuable from
the standpoint of evaluating the performance of the entire waste containment facility. This type
of performance monitoring also provides a valuable indication that the cover system is (or is not)
functioning as designed. However, if the actual performance of the cover system must be
quantified or definitively demonstrated, more direct monitoring methods will need to be used.
8.2.3 Drainage Layer Monitoring
As previously discussed in Section 1.5.3, conventional RCRA-type cover systems may require a
drainage layer installed between an overlying protection layer and underlying hydraulic barrier
(Figure 1-12), particularly on sideslopes. Drainage from this layer can be monitored: (i) as
DRAFT - DO NOT CITE OR QUOTE
8-4
-------
confirmation that the layer is functioning as intended; and (ii) to generate data on the water
balance for those components of the cover system above the drainage layer. Flow from the
drainage layer can only be quantified if the layer is designed to convey flow to not more than a
few discrete discharge points. At the discharge points, the flow rate can be monitored using a
flowmeter, tipping bucket, pore pressure transducer, or other means. If the drainage layer simply
daylights at the edge of the cover system and discharges as sheet flow to the surrounding area or
surface-water drainage structure, such as shown in Figure 2-5(a), quantitative monitoring will
not be possible. The need for monitorable discharge points results in a trade-off because, while
it is beneficial to collect monitoring data, construction of the discharge points may complicate
the design for some projects where simply daylighting the drainage layer would otherwise
suffice.
Currently, drainage layer monitoring is not routinely performed; it is usually only conducted for
a cover system test plot as part of a water balance assessment.
8.2.4 Lysimeter Monitoring
Lysimeters have long been used for agricultural and hydrologic studies to collect deep drainage
or percolation data or to estimate recharge. Lysimeters have been used to monitor percolation
through cover systems with hydraulic, ET, and capillary barriers. The most common approach is
to use a collection lysimeter, also called a pan lysimeter or drainage lysimeter. Other types of
lysimeters, including monolithic lysimeters, weighing lysimeters, and suction lysimeters, have
been used for various types of research studies, but not specifically for evaluation of installed
cover systems. The principal advantage of collection lysimeters is that, when properly designed,
they provide a direct measure of soil-water flux. Lysimeters perform best when they are
installed during cover system construction. When installed after the fact, great care is needed to
assure that the boundary conditions (e.g., vegetation and soil properties) above and adjacent to
the lysimeter are similar to the characteristics found elsewhere in the cover system.
The use of a collection lysimeter for percolation monitoring of a cover system is illustrated in
Figure 8-1. A lysimeter of the type shown in Figure 8-1 is constructed with hydraulic barrier
(typically GM) beneath or within the soil profile to be monitored. The liner is shaped to contain
percolation and is typically backfilled with a granular (sand or gravel) or geosynthetic drainage
layer. A geosynthetic drainage material may be preferred to a granular drainage material
because it has lower storage capacity and faster response time than most granular drainage
materials. Also, when a granular drainage material is used, it can impact the boundary
conditions of the cover systems and create the capillary barrier effect described in Chapter 1.
Liquid collected in the lined lysimeter drains by gravity to a monitoring point, where the flow is
collected and periodically measured with a pore pressure transducer, float and a pulse generator,
tipping bucket, or other means. To date, few lysimeters have been installed beneath full-scale
cover systems; instead, they have been installed beneath cover system test plots. However, this
statistic
DRAFT - DO NOT CITE OR QUOTE
8-5
-------
Monitoring
Station ~\
-GM
'Lysimeter
Discharge
Pipe
Figure 8-1. Example of a Collection Lysimeter Used to Monitor Percolation.
is changing since collection lysimeters are being installed beneath full-scale cover systems as
part of the ACAP program, which was discussed previously in Section 3.4.3. Generally, the
larger
the lysimeter, the more representative the monitoring results of the performance of the cover
system as a whole.
Runoff
Collection Pipe
- Percolation
Collection Pipe
GM
Percolation
Collection Lysimeter
Figure 8-2. Collection Lysimeter and Runoff Collection Pipe Used to Monitor Percolation
and Runoff at the Omega Hills, Wisconsin Test Plots Described in Section
7.2.1.
As described by Bonaparte et al. (2002), it appears that the best way to document the field
performance of CCLs in cover systems is with the use of lysimeters installed at the base of the
cover system. Five case studies reporting on the use of lysimeters to monitor percolation
through cover systems (e.g., Dwyer, 1997, 1998, 2001; Melchior, 1997a,b and Melchior et al.,
1994; Montgomery and Parsons, 1989, 1990; Nyhan et al., 1997; Paige et al., 1996) were
described previously in Section 7.2. A sixth case study using a similar technique was described
in Section 4.3.4 (Lane, 1992; Khire, 1995; Khire et al., 1997, 1999). The test plot and lysimeter
DRAFT - DO NOT CITE OR QUOTE
8-6
-------
set-up for the third and sixth case studies are illustrated in Figures 8-2 and 8-3, respectively.
These figures also illustrate how surface runoff was monitored for the test plots. Other examples
of the use of lysimeters in test plots are given by Webb et al. (1997) and Gee et al. (1997).
HOPE Cutoff—i
Vegetated—:
Surface
Layer
Berm—,
PVCPipe—,
Nest of
Thermocouples —|
and
TDR Probes
Berm
GC
Drainage Layer
Overland Flow Tank
Figure 8-3. Collection Lysimeter and Runoff Collection Pipe Used to Monitor Percolation
and Runoff at the Live Oak, Georgia and Wenatchee, Washington Test Plots
Described in Section 7.2.4.
8.3 Soil Moisture and Matric Potential Monitoring
8.3.1 Overview
Soil moisture and matric potential measurements can be used to assess soil moisture or matric
potential content at discrete locations, changes in cover system water storage, and vertical
gradients in cover system soils. With careful calibration, the measured moisture contents can be
converted to matric potentials, and vice-versa, through the use of an acceptable soil-moisture
characteristic curve. Currently available techniques of assessing soil moisture content in cover
systems include neutron probes, time domain reflectrometry (TDR) probes, and frequency
domain reflectometry (FDR) probes. Methods of measuring soil matric potential include
DRAFT - DO NOT CITE OR QUOTE
3-7
-------
tensiometers, electrical resistance sensors, thermocouple psychrometers, and heat dissipation
sensors. TDR and FDR probes have also been used to measure matric potential when they have
been combined with a matrix material whose water retention function has previously been
determined. With these modified sensors, the matrix material around the TDR or FDR probes
comes into equilibrium with the surrounding soil, and the water content (and indirectly the
matric potential) of the matrix material is measured with the probes. These modified probes will
not be discussed further.
All of the soil moisture and matric potential monitoring methods listed above are non-
destructive, in-direct techniques. With the exception of thermocouple psychrometers, good
contact between the sensor and the soil (or borehole casing for neutron and FDR probes) is
required to obtain accurate measurements. This is especially critical for sensors that measure
matric potential and rely on good hydraulic contact with the soil to establish thermodynamic
equilibrium. Good hydraulic contact may be hard to attain in very coarse soils, such as gravel,
and in shrink-swell clays. Except for the neutron probe, all of the sensors can be fully
automated.
Soil moisture and matric potential measurements may also be made directly on soil samples
excavated from the cover system. While this latter method is reliable for determining soil
moisture content or matric potential, it involves destructive sampling (i.e., damage) of the cover
system soil and the inherent problem of sample variability associated with the destructive
sampling protocol.
8.3.2 Neutron Probes
The neutron probe, when calibrated, can yield very good indirect measurement of soil moisture
content. The probe is inserted into a cased access borehole, orientated in any direction, where
readings are taken at various locations (Figure 8-4). The casing material is generally aluminum
or PVC piping. The principle of operation is based upon the neutron thermalization process,
wherein a radioactive source emits high-energy neutrons, with an energy of about 5 MeV, into
the soil. These neutrons are then reduced to a lower energy state upon colliding with hydrogen
atoms associated with soil water (Gardner, 1987). After an average of 19 collisions, the neutrons
cease to lose further energy and are said to be "thermal" neutrons with an energy of
approximately 0.025 MeV. Higher molecular weight elements, such as oxygen, also slow the
neutrons, but far fewer collisions are required with hydrogen to slow the reaction to thermal
energy levels. The source of the high-energy neutrons in most commercially available neutron
probes is a radioactive americium and beryllium mix. The americium emits an alpha particle
that bombards the beryllium atoms, which, in turn, emit a neutron. The fast neutrons are emitted
approximately radially from the source and form a sphere around the source within which the
neutrons are attenuated. The size of this spherical influence varies inversely with the moisture
content. The sphere is about 0.7 m for dry soil and about 0.16m for saturated soil; sphere
diameter is unaffected by the strength of the radioactive source (Gardner, 1987). The number of
pulses counted by the probe detector is proportional to the number of thermal neutrons
encountered. A calibration curve can be developed to correlate count rate with soil moisture
content.
DRAFT - DO NOT CITE OR QUOTE
-------
While the probe's manufacturer usually supplies calibration curves, calibration for each
application is recommended. A calibration involves taking multiple readings in a given soil
against a range of gravimetrically-determined moisture contents. Soil heterogeneity and organic
matter can have adverse affects on accuracy of neutron probe readings. Also, extraneous
Control Unit
Probe Housing
Neutron Source
and Detector
Figure 8-4. Neutron Probe Installed in a Vertical Cased Borehole.
hydrogen atoms not associated with water can also impact probe accuracy. Potential sources of
the extraneous hydrogen atoms include hydrocarbons, methane gas, hydrous minerals (e.g.,
gypsum), hydrogen-bearing minerals (e.g., kaolinite, illite, and montmorillonite), and organic
matter in the soil. Irregularities in the borehole casing or contact with the soil around the
perimeter of the borehole can also produce error in moisture content values obtained. A
disadvantage to the use of a neutron probe is the fact that a radioactive source is present, thereby
posing a potential hazard for the operator as well as imposing difficulty in its use and
maintenance (i.e., regulatory constrains). In addition, because of the regulatory constraints for
using the radioactive source, this method cannot be automated.
DRAFT - DO NOT CITE OR QUOTE
8-9
-------
In cover system monitoring applications, neutron probes are typically placed into access tubes in
the cover system, and water content measurements are made at discrete locations at discrete time
intervals. Neutron probes have been used to monitor soil moisture content in cover systems at a
number of sites (e.g., Montgomery and Parsons, 1989, 1990; Nyhan et al., 1990; Payer et al.,
1992; Anderson et al., 1993; Schultz et al., 1995; Paige et al., 1996).
8.3.3 Time Domain Reflectometry
The process of sending electromagnetic pulses through a conductor and observing the reflected
waveform is called time domain reflectometry (TDR). When monitoring soil moisture, TDR
equipment generally consists of a cable tester or a specially designed commercial TDR unit,
coaxial cable, and a stainless steel probe (Figure 8-5). The type of material surrounding the
conductor (i.e, cable and probe) influences the waveform traveling down the conductor. The
waveform is reflected differently when it reaches the start of the probe and the end of the probe.
Figure 8-5. TDR Probe and Coaxial Cable.
DRAFT - DO NOT CITE OR QUOTE
8-10
-------
The time-of-travel along the conductor is dependent on the dielectric constant of the surrounding
medium (e.g., the sheath around the cable or the soil around the probe). If the dielectric constant
of the medium surrounding the conductor is high, the electronic signal propagates more slowly.
Because the dielectric constant of water is much higher than most materials, a signal within a
wet or moist medium propagates slower than in the same medium when dry. The dielectric
constant of water is about 80, whereas the dielectric constant of dry soil is typically in the range
of about 3 to 5. Ionic conductivity affects the amplitude of the signal but not the propagation
time. Thus, soil moisture content around the probe can be assessed by a pre-determined
correlation between time-of-travel along the probe (obtained from analysis of the reflected
waveform) and soil moisture content. A generic calibration equation developed by Topp et al.
(1980) is sometimes used. However, the probes should be calibrated for their specific
application (e.g., soil texture and density and cable length) to yield accurate soil moisture
measurements (Lopez and Dwyer, 1997).
The accuracy of TDR for soil moisture measurements is relatively good for many soil types and,
according to Schofield et al. (1994), about the same as that for neutron attenuation. A
disadvantage of TDR is the fact that accuracy decreases with increased cable length between the
probe and the cable tester; generally a maximum range of about 60 m is recommended. In
addition, soils with a high moisture content and a high electrical conductivity rapidly attenuate
the electrical pulse before it is reflected back. If the attenuation is great enough there will be no
return signal and the probe cannot be used. However, probes can be coated to reduce signal
attenuation.
TDR Cable
6 in.
36 in.
Figure 8-6. Example of Horizontally Orientated TDR Probes in a Cover System.
DRAFT - DO NOT CITE OR QUOTE
8-11
-------
Probes may be installed during or after construction. They can be installed in any direction;
however, when installed after construction, they are usually inserted vertically. When installed
in this fashion, care should be taken to minimize the soil disturbance around the probe such that
the probe fits snuggly in the soil. There have been cases where a space formed between the
probe and soil during installation such that water was able to infiltrate into the space and short-
circuit the cover system during heavy rainfall events. Consequently, the water content
measurements at the probe were not representative of the surrounding cover system soils.
Recent developments have attempted to minimize the cable length problem and reduce the cost
of the TDR system. The latest development is a probe that does not require a cable tester or
TDR unit but rather connects directly to a data logger. Calibration similar to the traditional TDR
system is required for best results. The probe consists of two stainless steel rods connected to a
printed circuit board. A five-conductor cable is connected to the circuit board to supply power,
activate the probe, and monitor pulse output. The circuit board is potted in an epoxy block.
TDR has been used to monitor soil moisture content in cover systems at a number of sites (e.g.,
Dwyer, 1997, 1998, 2001; Kavazanjian, 2000; Khire, 1995; Khire et al., 1997,1999; Lane et al.
1992; Montgomery and Parsons, 1989, 1990; Nyhan et al., 1997). The use of TDR for soil
moisture content monitoring is illustrated in Figure 8-6.
8.3.4 Frequency Domain Reflectometry
Frequency domain reflectometry (FDR) methods of soil moisture content measurements are also
known as radio frequency (RF) capacitance techniques. These techniques actually measure soil
capacitance. The probe contains a pair of electrodes and the soil serves as the dielectric medium
completing a capacitance circuit comprising part of a feedback loop of a high frequency
transistor oscillator. As high frequency radio waves (about 150 MHz) are pulsed through the
capacitance circuitry, a natural resonant frequency dependent upon the soil capacitance is
established. The soil capacitance is related to the dielectric constant by the geometry of the
electric field established around the electrodes. Either the natural resonant frequency or the
frequency shift between the emitted and received frequencies is recorded.
The FDR probe is often used in an access tube (cased borehole) similar to the neutron probe for
measuring soil moisture content at various depths. In this application, it is important that the
access tube be sized to provide a snug fit around the probe, thereby minimizing annular air gaps
that greatly affect the travel of the electronic signal into the soil. Installation of the access tube
also requires special attention to ensure complete soil contact with the casing since annular air
gaps or soil cracks around the outside of the tube also produce erroneously-low readings.
Though the FDR probe manufacturer may provide calibration curves, it is important that the
probe be calibrated with the site-specific soil. With proper calibration and use, the accuracy of
the FDR method for measuring soil moisture content is good.
DRAFT - DO NOT CITE OR QUOTE
8-12
-------
8.3.5 Tensiometers
A tensiometer measures soil matric potential values between 0 and approximately -90 kPa. The
range of measurement is limited by the cavitation of water, which occurs at matric potentials less
than -100 kPa. A tensiometer commonly consists of a high air entry, porous ceramic cup
connected to a pressure measuring device through a rigid plastic tube (Figure 8-7). Plastic is the
preferred material for the tube because of its non-corrosive nature and lower heat conduction
properties. The tube is sealed at the top with a removable cap allowing the tensiometer to be
filled with deaired water and accumulated air to be purged. A Bourdon gauge, manometer, or
Water Reservoir-
Water Level-
Plastic Tube -\
I
Porous Cup-
f
Removable cap
Vacuum Gage
Figure 8-7. Tensiometer.
pressure transducer is attached to the upper portion of the water-filled tube to measure the
negative pressure of the water in the tensiometer. The matric potential of the soil is equal to this
negative pressure plus a pressure correction that accounts for the elevation potential of the water
column in the tensiometer.
When the tensiometer is inserted into the soil, the soil absorbs water from the tensiometer and as
this occurs the water pressure in the tensiometer decreases until the tensiometer fluid pressure is
in equilibrium with soil water matric potential outside the cup. Tensiometers are limited to moist
soils.
DRAFT - DO NOT CITE OR QUOTE
1 1
-13
-------
8.3.6 Electrical Resistance Sensors
Electrical resistance sensors have been used for over 60 years in agricultural applications
(Bouyoucos and Mick, 1940). They consist of electrodes embedded in a gypsum, nylon, or
fiberglass porous material that equilibrates with the surrounding soil. During equilibrium, water
and solutes exchange between the sensor and the soil; therefore, the matric potential of the
sensor is the same as that of the soil after equilibrium. Although electrical resistance varies
primarily with water content, the equilibrium between the sensor and the soil is a matric potential
rather than a water content equilibrium. These dual relationships result in a hysteretic
relationship between the sensor's electrical resistance and matric potential. In practice, the
sensors are more often calibrated to soil water content than to matric potential.
The electrodes in electrical resistance sensors have leads connected to a Wheatstone bridge to
measure resistance. When the sensor is placed in firm contact with the soil, water flows into or
out of the sensor until equilibrium is established. As the moisture content of the resistance block
decreases, the electrical conductivity of the block decreases and the electrical resistivity of the
block increases. Ohmmeters are used to measure resistance. The upper measurement range of
the sensors is controlled by the air entry pressure of the sensor matrix material, and the lower
limit depends on the range in smaller pore sizes of the sensor matrix. For gypsum blocks, the
upper limit is approximately -30 kPa (Bourget et al., 1958) and the lower limit is approximately
-1000 kPa (Tanner et al., 1952; Bourget et al., 1958). Additional discussion of gypsum blocks
and fiberglass moisture sensors are given below.
Daniel et al. (1992) described gypsum blocks as prismatic or cylindrical blocks of gypsum that
change electrical resistance when they change moisture content. The gypsum block is placed in
the soil and the gypsum either takes in water from or gives up water to the surrounding soil until
thermodynamic equilibrium is established. The electrical resistance of gypsum varies with
moisture content: the higher the moisture content, the higher the electrical conductivity and,
hence, the lower the electrical resistance. Because gypsum is partly soluble in water, it gives the
sensor a buffering capacity that makes it insensitive to soil electrolyte concentrations less than
about 300 ppm (2 mmhos/cm). However, for salt concentrations greater than 5,000 ppm in the
surrounding soil, the electrolyte concentration in, and electrical resistance of, gypsum blocks can
be affected. As a result of their solubility, gypsum blocks placed in wet soils tend to
disintegrate. However, resins may be added to gypsum to improve their longevity. It has been
reported that gypsum blocks may function for more than 5 years in dry soils but as little as 3
months in wet soils.
Daniel et al. (1992) report that fiberglass sensors work in much the same way as gypsum blocks;
however, they don't have the buffering capacity that is provided by the dissolving gypsum. A
porous fiberglass cloth is placed in the soil; the fiberglass gains or loses water until
thermodynamic equilibrium is reached. A temperature-measuring probe may be a part of the
unit.
Both gypsum blocks and fiberglass sensors were used to monitor the performance of the GCL
test plots described previously in Section 7.4.5. The shapes and dimensions of the sensors used
DRAFT - DO NOT CITE OR QUOTE
8-14
-------
in the test plots are shown in Figure 8-8, and a typical placement of the sensors within a test plot
cover system cross section is indicated in Figure 8-9. As indicated by Figure 8-9, the gypsum
blocks were placed in the subgrade beneath the cover system geosynthetics. The fiberglass
sensors were placed at the subgrade/GCL and GCL/GM interfaces.
(a) Gypsum Block
(b) Fiberglass Moisture Sensor
40mm
40 mm
Figure 8-8. Dimensions of Electrical Resistivity Sensors Used in GCL Test Plot
Described in Section 7.4.5.
GC
'^tiiW^^P^W^I
"S^ v" N^ V1 ^>
Soil Cover
w
GCL—^ I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I i 11
i h- — hiuergia
... '•:•:;•.•.:••]•.•'.:•::•.•.'••• --1 .•.':•"'•.'•/.
\— Gypsum Block
Figure 8-9. Layout of Electrical Resistivity Sensors Used in GCL Test Plot Described in
Section 7.4.5.
DRAFT - DO NOT CITE OR QUOTE
8-15
-------
8.3.7 Thermocouple Psychrometers
A psychrometer infers the matric potential of the liquid phase of a soil from measurements
within the vapor phase that is in equilibrium with the sample. It measures the relative humidity
within a soil system as the difference between a dry bulb (non-evaporating) temperature and a
wet bulb (evaporating) temperature. The primary difficultly with this technique is that the
relative humidity in the soil gas phase changes only a small amount within the typical range of
interest. For example, at 25 °C, a water potential of-1.5 MPa (wilting point) corresponds with a
relative humidity of about 0.99, and a water potential of-8 MPa (lower limit of extraction for
many desert plants) corresponds with a relative humidity of 0.94. Thus, practically all
measurements of interest to most cover system studies lie in a narrow relative humidity range
between 0.94 and 1.0. Thermocouple psychrometers are typically used to monitor matric
potentials in the range of-8MPa to -30 kPa.
The majority of psychrometers used in the field utilize the Spanner design. This design is
composed of a thermocouple, a reference electrode, a heat sink, a protective porous ceramic bulb
or wire mesh screen, and a recorder. The technique is based on measuring the temperature of a
water droplet or wet surface using thermocouple junctions. Calibration curves are developed by
immersing the unit in a series of sodium or potassium chloride solutions of known concentration
(generally 0.1, 0.3, 0.5, 0.8, and 1.0 molar (Morrison, 1983)) at specified temperatures. The
calibration curves are used to compute the in-situ soil-water potential from the measured field
output voltage. Problems sometimes encountered with psychrometers are that their calibration
can change over time due to corrosion (Daniel et al., 1981) and/or microbial growth on the
thermocouple wires (Merrill andRawlins, 1972).
8.3.8 Heat Dissipation Sensors
Heat dissipation sensors, also called thermal conductivity sensors (Fredlund, 1992) or matric
potential sensors, rely on the relationship between the heat dissipation of a ceramic matrix in
contact with soil and the matric potential of the soil. These sensors also have a relatively long
history of use in agricultural studies. The sensor consists of a heater and a temperature sensor in
a ceramic matrix (Figure 8-10). A current is applied to the heater and the temperature of the
sensor is measured at certain time intervals, typically at 1 and 20 s after the initiation of heating.
The change in temperature (i.e., the heat dissipation) is controlled by the water content of the
ceramic matrix because water conducts heat much more readily than air (i.e., thermal
conductivity increases with water content). The measured temperature increase represents the
heat that is not dissipated. The temperature increase is calibrated to sensor matric potential.
The upper measurement range of the sensor is controlled by the air entry pressure of the sensor
matrix material, which is generally about -lOkPa. The lower limit is generally considered to be
about -1 MPa (Reece, 1996). The sensitivity of the heat dissipation sensors decreases as soils
dry below -1 MPa.
Heat dissipation sensors have been used to monitor soil matric potential in cover systems at a
number of sites, including the ACAP test sites.
DRAFT - DO NOT CITE OR QUOTE
8-16
-------
Figure 8-10. Heat Dissipation Sensor.
8.4 Gas Emissions Monitoring
Gas emissions measurements can be used to assess the performance of cover systems and gas
control systems. Gas emissions are a common concern for MSW landfills or CERCLA sites that
contain MSW. Landfill methane emissions measured at MSW landfill sites and reported in the
literature have ranged from about 0.003 to 3,000 g/m2/d (Bogner and Scott, 1997). In general,
the higher rates were associated with landfills that did not have gas recovery and that were
covered with dry soils without a GM barrier. For example, at the Olinda MSW Landfill in
Southern California, which is covered by a sandy silt soil layer, measured emission rates were
greater than 1,000 g/m2/d prior to installation of a gas collection system. After a gas collection
system was installed, measured gas flux rates were less than 10 g/m2/d. The flux rates were still
lower (less than 0.01 g/m2/d) in the area of the landfill with a gas recovery system and covered
with a clayey silt layer. Given this wide range of emissions, it is appropriate at many MSW sites
DRAFT - DO NOT CITE OR QUOTE
8-17
-------
and CERCLA sites that contain MSW to divide the sites into areas with different surface
characteristics, moisture regimes, and gas control strategies and obtain order-of-magnitude
estimates of fluxes from these areas for the purposes of assessing emissions. For HW landfills or
waste piles that began operation after EPA passed its Land Disposal Restrictions, emissions are
generally at much lower rates than recently filled MSW landfills and it may be possible to install
a passive venting system.
Landfill gas emission rates can be measured indirectly or directly. Subsurface vertical methane
gradients calculated using Pick's First Law (i.e., assuming diffusive transport only) and
measured concentrations at gas probes at various depths have been used to estimate gas
emissions. This indirect method typically results in higher estimated fluxes than those measured
using a direct chamber technique (e.g. a flux chamber) (Rolston, 1986). However, the indirect
method is often useful as an independent check on emission values obtained using a flux
chamber (Bogner and Scott, 1997). The most common direct methods for monitoring landfill
gas emissions are vent sampling and the flux chamber techniques. The most common means of
evaluating gas emissions is by using indirect methods (i.e., back-calculating emissions from the
source based on a measured concentration). One method of indirect monitoring (described in
EPA, 1992) involves concentration profile sampling. The sampling device is placed at the cover
system with sampling probes spaced at different intervals. The concentration, wind speed, and
temperature are measured at each of the probe heights to generate profiles for each. This
technique does not work when quiescent or unstable wind conditions exist, such as shifting of
direction. The site must be relatively homogeneous; the technique will not work if emissions or
waste composition vary with respect to locations. In all cases, gas sampling must be conducted
over a period of time, since gas emission rates are not constants.
The transect technique, performed with a device that used both a vertical and horizontal array of
sampling probes placed downwind of the source in the plume centerline, has also been used.
Background measurements are also made upwind of the source to correct for the contribution
from other sources. The device also has instruments to measure wind speed, wind direction, and
temperature. The measured concentrations are spatially integrated and a Gaussian dispersion
model is used to back-calculate the emission rate from the source that would be needed to give
the measured concentration.
Instantaneous Surface Monitoring (ISM), Integrated Surface Sampling (ISS), and flux chamber
techniques (Cooper and Bier, 1997; Lu and Kunz, 1981). Each of these methods is described
below. For all of these methods, monitoring is generally not conducted within 72 hours
following a precipitation event to allow the cover soil to drain (the trapped water in the soil
impedes emissions). Gas emissions can also be measured by vent sampling, which requires the
volumetric rate of flow be measured.
With ISM, a portable flame ionization detector (FID) is used to measure the instantaneous
concentration of total organic compounds (TOCs) (as methane) along transects or grids
established at the landfill surface. This method does not measure flux, but can be used to divide
the site into areas with different emission rates. The specific emission rates of these areas can
then be evaluated using a flux chamber.
DRAFT - DO NOT CITE OR QUOTE
8-18
-------
ISS uses a grid-based method to collect samples of the surface gases. Within each grid square, a
8 to 10 L sample of gas is continuously collected from about 50 to 75 mm above the soil cover
surface over 25 minutes. Thus, the method provides an average constituent concentration, but
not flux, in a grid square. The gas samples can be analyzed in the field or laboratory.
Flux chamber methods have been used at landfill sites since at least the late 1970's (e.g., at the
Fresh Kills Landfill, New York (Lu and Kunz, 1981). They involve enclosing a known volume
of atmosphere above a known soil surface area and obtaining a direct, though spatially limited,
measurement of emission rate. Flux chambers represent a compromise as they may influence
flow fields, temperature, and concentrations at the soil/atmosphere interface. However, they
have significant advantages if they are operated over short time periods and minimize
disturbance. Also, unlike the ISM and ISS methods, the flux chamber method can be used to
monitor emissions in high winds; the ISM and ISS methods should generally not be performed
when the average wind speed exceeds 16 kph to avoid dilution of the emitted gas by air (Cooper
and Bier, 1997). The sensitivity of the flux chamber method can be adjusted by varying the flux
chamber volume. They are good for measurement over a 1 to 10 m2 scale, but are typically less
than 1 m2 with a volume less than 20 L.
8.5 Settlement Monitoring
Post-closure settlement monitoring should consider both total and differential landfill
settlements. In general, differential settlements are of most concern because they may induce
unacceptable tensile stress and strain in one or more cover system components and they may
cause cover system slopes to change or reverse grade. As previously discussed in Section 6.4,
cover system settlement can be considered to have one of three sources: (i) settlement of
foundation soil; (ii) settlement due to overall waste mass compressibility; and (iii) settlement due
to localized mechanisms in the waste. When monitoring cover system settlements, the sources
of the settlements are not differentiated; rather, the total settlement at any point due to all of
these sources is measured. The measured settlements are then evaluated to assess the effect of
the settlements on the cover system components and slopes. For example, most compacted clays
exhibit failure at extensional strains of 0.5% or less, as discussed in Section 6.4.5.
Procedures for monitoring total settlements of the cover system surface include:
aerial surveys, which are generally limited to a vertical accuracy of about 100 to 200 mm
with good ground control, and are often more expensive than ground surveys depending
on the size of the survey area; the accuracy of aerial surveys may be impacted by a
number of things including time of day, angle of sun, and cloud and ground cover; they
also require a certain amount of field surveying for ground-truthing and targeting;
conventional instrumental ground surveys of settlement monuments installed on the
cover system, which can achieve high precision (vertical accuracy to within less than 1
mm); and
global positioning system (GPS) surveys performed using hand operated equipment; the
precision and cost of GPS surveys are a function of the specific equipment used; if
DRAFT - DO NOT CITE OR QUOTE
8-19
-------
significant vegetation is present, GPS may be less reliable since it may be difficult to
receive satellite signals.
In addition to total settlements of the cover system surface, settlement of the components within
the cover system are sometimes monitored. For example, the settlements of the cover system
components for a low-level radioactive waste landfill are being monitored by settlement plates
and ground penetrating radar (GPR). The settlement plates were installed above the drainage
layer during cover system construction to verify that sufficient drainage layer slope is being
maintained. The GPR targets were installed at different locations within the protection layer
(Figure 8-11). Both the settlement plates and the GPR targets are periodically surveyed.
Figure 8-11. Placement of GPR Target on Top of Drainage Layer (and at the Bottom of
the Protection Layer) During Cover System Construction.
Settlement monuments can be installed on cover system slopes to monitor for downslope creep
or instability. This type of monitoring may not be necessary for cover systems designed to
conventional factors of safety (as defined in Chapter 6 of this document). However, for
situations where lower factors of safety are utilized, slope monitoring is advisable. Slope
monitoring should also be considered for final cover systems in seismic impact zones where the
DRAFT - DO NOT CITE OR QUOTE
8-20
-------
cover system is designed to yield (undergo permanent seismic displacement) during the design
earthquake event. Slope inclinometers can also be used to monitor for slope movements.
Differential settlement monitoring may be identified through aerial or ground survey techniques
if the differential settlement feature is large enough to be captured by the resolution of the survey
technique used. Area-wide surface depressions less than 300 to 600 mm in depth are unlikely to
be identified through aerial survey. Likewise, highly localized raveling (fines moving into larger
voids) or sinkhole features are likely to go undetected in aerial surveys. These same features
would be missed in ground surveys where it would be unusual, for example, to install settlement
monuments on a survey grid with a grid dimension smaller than about 30m. The most reliable
means for identifying localized differential settlements is to perform periodic visual surveys
across the entire landfill surface. This type of survey should ideally be performed immediately
after a rainstorm when puddles and ponded water would provide evidence of surface
depressions. At the same time, the cover system can be inspected for evidence of other types of
differential settlement features such as sinkholes, gullies, or raveling conditions. Also,
experience indicates that contrasts develop in surface vegetation in and around depressions, since
the cover soil in the depression tends to stay wetter than elsewhere. Thus, contrasts in cover
vegetation color and health can be used to identify locations where surface depressions might
exist.
As pointed out in EPA (1991), subsidence depressions should be remediated below the level of
the hydraulic barrier to avoid long-term acceleration of the subsidence due to a "roof ponding"
mechanism. Roof ponding refers to the common structural problem in flat roof systems where
ponding water causes the roof rafters to deflect, thus allowing more water to pond, causing more
deflection, and so on. This mechanism continues until the roof collapses. In addition, ponding
above a portion of the hydraulic barrier increases the potential for percolation through the barrier
within the ponded area. Remediation requires removing the cover system in the region of
subsidence, backfilling the depression with fill, and then reconstructing the cover system in the
repaired area. To minimize the potential for continuing settlement, the use of engineering
measures such as geosynthetic reinforcement or separation layers, lightweight fill, vibratory
compaction of backfill (to help fill ravel features and voids), etc. should be considered.
DRAFT - DO NOT CITE OR QUOTE
8-21
-------
Chapter 9
Post-Closure Maintenance and Site End Use
9.1 Introduction
After a cover system has been constructed, it must be monitored and maintained for some
timeframe (i.e., the post-closure period). As discussed in Sections 1.2.6 and 8.1, post closure
maintenance must be conducted as long as the waste poses a threat to human health and the
environment. The post-closure period of 30 years given in RCRA regulations has generally been
considered by EPA to be the minimum timeframe for performance monitoring and maintenance
for MSW and HW facilities. For CERCLA facilities, the minimum timeframe for cover system
maintenance and monitoring is also often assumed to be 30 years, and the EPA is required to
evaluate the performance of the cover system at least once every five years to assure that human
health and the environment are being protected by the implemented remedy.
Regulatory requirements for post-closure maintenance of MSW landfill cover systems are
contained in 40 CFR §258.61 (a)(l):
"(a) Following closure ofeachMSWLF unit, the owner or operator must conduct post-
closure care. Post-closure care must be conducted for 30 years, except as provided under
paragraph (b) of this section, and consist of at least the following:
(1) Maintaining the integrity and effectiveness of any final cover, including making repairs
to the cover as necessary to correct the effects of settlement, subsidence, erosion, or other
events, and preventing run-on and run-off from eroding or otherwise damaging the final
cover."
For MSW landfills, 40 CFR §258.61 (b) provides the following flexibility with respect to the
length of the post-closure period:
"(b) The length of the post-closure care period may be:
(1) Decreased by the Director of an approved State if the owner or operator demonstrates
that the reduced period is sufficient to protect human health and the environment and this
demonstration is approved by the Director of an approved State; or
(2) Increased by the Director of an approved State if the Director of an approved State
determines that the lengthened period is necessary to protect human health and the
environment."
Analogous requirements for HW landfills are contained in 40 CFR §264.310 (b)(l) and (5).
Regulations for MSW landfills presented in 40 CFR §258.61(c) and regulations for hazardous
waste facilities presented in 40 CFR §264.118 require facility owners or operators to prepare a
written post-closure plan that includes a description of the post-closure maintenance activities
and the frequency of such activities. The purpose of these activities is to ensure the integrity of
the cover system and functionality of any monitoring equipment. Maintenance activities include
those conducted in response to observations made during periodic inspections and monitoring
DRAFT - DO NOT CITE OR QUOTE
9-1
-------
and scheduled routine activities, such as pump maintenance or replacement. An example of a
post-closure inspection, monitoring and maintenance schedule is presented in Table 9-1. An
example of a post-closure inspection form, used by the U.S. Army Corps of Engineers, is
presented in Table 9-2. This table can be used to document the condition of a landfill cover and
identify any required post-closure maintenance activities. In addition to regularly scheduled
inspections, a thorough inspection of the cover system should be conducted after major storm
events.
The maintenance (and monitoring) activities to be conducted at a closed waste containment
facility or remediation site depend on the end use of the site. For example, as discussed in
Section 9.3.5, when a mountain bike challenge course was constructed on top of a cover system,
routine cover system maintenance included repairing ruts made by the bike tires. It is
recommended that personnel conducting the maintenance activities be familiar with the function
of the cover system, rather than only familiar with the site end use (e.g., sports facility). If
maintenance is not correctly performed, cover system or monitoring system integrity may be
impaired.
Table 9-1. Example of waste containment facility or remediation site monitoring and
maintenance schedule.
Component
Cover System Vegetation
Cover System Erosion
Cover System Intrusion
Cover System Subsidence
Cover System Slope Stability
Cover System Drainage Outlets
Cover System Grades (Survey)
Gas Extraction System
Surface-Water Management System
Leachate Collection and Removal System/
Leak Detection System
Perimeter Security (fence, gate, locks)
Access Roads
Groundwater Monitoring System
Gas Monitoring System
Survey Monuments
Post-Earthquake Condition of all
Systems/Structures
Inspection and Monitoring Methods2
Frequency1
Monthly Visual
Monthly and After Major Storms Visual
Monthly Visual
Quarterly Visual
Quarterly Visual
Quarterly Visual
Every 5 Years Survey/GPS
Monthly System Check
Quarterly and After Major Storms Visual
Monthly System Check
Quarterly Visual
Quarterly Visual/RT/PC
Quarterly System Check
Quarterly System Check
Annually for First 5 Years, at 5 Year Survey
Intervals Thereafter
After Earthquakes All Above
Vrequency of inspection and monitoring may be reduced (or increased) based on observed conditions during the
post-closure period.
GPS = global positioning system; RT = rut depth for unpaved roads; and PC = pavement cracking for paved roads.
DRAFT - DO NOT CITE OR QUOTE
9-2
-------
This chapter discusses cover system maintenance and site end use. Other types of post-closure
maintenance activities typically associated with waste containment facilities or remediation sites
are not addressed herein. These include maintenance of leachate collection and removal
systems, leak detection systems, groundwater monitoring systems, and gas management and
monitoring systems. The condition of these systems must be monitored during the post-closure
period to assure adequate performance of the site in the long term and to comply with various
regulatory requirements.
DRAFT - DO NOT CITE OR QUOTE
9-3
-------
Table 9-2. Example of post-closure monitoring form used by U.S. Army Corps of
Engineers for CERCLA sites.
Site Name:
CERCLIS ID:
State:
Corps Construction District:
EPA Region:
Inspection Team: Attach Roster
ITEM
Date of Inspection:
Weather:
Temperature:
Corps Design District:
Site Map: Attach
Mote: Indicate the location of any deficiency noted
below on the site map
REMARKS
COVER SYSTEM SURFACE
1 . SETTLEMENT (LOW SPOTS) Yes ( ) No ( )
Areal Extent:
Depth:
2. CRACKS Yes ( ) No ( )
_ength:
Width:
Depth:
3. EROSION Yes( ) No ( )
Areal Extent:
Depth:
4. HOLES Yes ( ) No ( )
Areal Extent:
Depth:
Suspected Cause (Rodent or Other):
5. VEGETATIVE COVER Yes ( ) No ( )
Grass: Yes No
Condition:
Trees/Shrubs Yes ( ) No ( )
Size:
6. ARMORED COVER Yes ( ) No ( )
Vlaterial Type:
Condition:
7. BULGES Yes ( ) No ( )
Areal Extent:
Height:
Suspected Cause (gas pressure or other):
8. WET AREAS Yes ( ) No ( )
Donding: Yes ( ) No ( )
Areal Extent:
Seeps: Yes ( ) No ( )
Areal Extent:
Estimated Flow Rate:
Soft Subgrade: Yes ( ) No ( )
Areal Extent:
9. SLOPE INSTABILITY Yes ( ) No ( )
Slides: Yes ( ) No ( )
Areal Extent:
Probable Slide Interface:
Suspected Cause:
Exposed Cover Components:
DRAFT - DO NOT CITE OR QUOTE
9-4
-------
Table 9-2. Example of post-closure monitoring form used by U.S. Army Corps of
Engineers for CERCLA sites (cont).
BENCHES
1. FLOW BYPASS BENCHES Yes ( ) No ( )
Description of problem:
2. BENCH BREACHED Yes ( ) No ( )
Description of problem:
3. BENCH OVERTOPPED Yes ( ) No ( )
Description of problem:
LETDOWN CHANNELS
1. SETTLEMENT Yes ( ) No ( )
Areal Extent:
Depth:
2. MATERIAL DEGRADATION Yes ( ) No ( )
Material Type:
Areal Extent:
Degree of Degradation:
3. EROSION Yes( ) No ( )
Areal Extent:
Depth:
4. UNDERCUTTING Yes ( ) No ( )
Areal Extent:
Depth:
5. OBSTRUCTIONS Yes ( ) No ( )
Type:
Areal Extent:
Size:
6. SLOPE INSTABILITY Yes ( ) No ( )
Type:
Areal Extent:
COVER PENETRATIONS
1. GAS VENTS Yes( ) No ( )
Active ( ) Passive ( )
=unctioning: Yes ( ) No ( )
Condition:
Routinely Sampled: Yes ( ) No ( )
2. GAS MONITORING PROBES Yes ( ) No ( )
Functioning: Yes ( ) No ( )
Condition:
Routinely Sampled: Yes ( ) No ( )
3. MONITORING WELLS Yes ( ) No ( )
Functioning: Yes ( ) No ( )
Condition:
Routinely Sampled: Yes ( ) No ( )
4. LEACHATE EXTRACTION WELLS Yes ( ) No ( )
=unctioning: Yes ( ) No ( )
Condition:
Routinely Sampled: Yes ( ) No ( )
5. SETTLEMENT MONUMENTS Yes ( ) No ( )
Located: Yes ( ) No ( )
Condition:
Routinely Surveyed: Yes ( ) No ( )
DRAFT - DO NOT CITE OR QUOTE
9-5
-------
Table 9-2. Example of post-closure monitoring form used by U.S. Army Corps of
Engineers for CERCLA sites (cont).
COVER DRAINAGE LAYER
1. OUTLET PIPES Yes ( ) No ( )
Functioning: Yes ( ) No ( )
Condition:
2. OUTLET ROCK Yes ( ) No ( )
Functioning: Yes ( ) No ( )
Condition:
DETENTION/SEDIMENTATION PONDS
1. SILTATION Yes( ) No ( )
Areal Extent:
Depth:
2. EROSION Yes( ) No ( )
Areal Extent:
Depth:
3. OUTLET WORKS Yes ( ) No ( )
Functioning: Yes ( ) No ( )
Condition:
4. Embankment Yes ( ) No ( )
Functioning: Yes No
Condition:
RETAINING WALLS
1. DEFORMATIONS Yes ( ) No ( )
Horizontal Displacement:
Vertical Displacement:
Rotational Displacement:
2. DEGRADATION Yes ( ) No ( )
Description of damage:
VERTICAL BARRIER WALLS
1. SETTLEMENT Yes ( ) No ( )
Areal Extent:
Depth:
2. PERFORMANCE MONITORING Yes ( ) No ( )
Type of Monitoring:
Frequency:
Evidence of Breaching: Yes ( ) No ( )
GROUNDWATER SYSTEMS
TYPE OF SYSTEM: Containment ( ) Treatment ( )
Functioning: Yes ( ) No ( )
Condition:
Routinely Monitored: Yes ( ) No ( )
DRAFT - DO NOT CITE OR QUOTE
9-6
-------
Table 9-2. Example of post-closure monitoring form used by U.S. Army Corps of
Engineers for CERCLA sites (cont).
PERIMETER DITCHES/OFF-SITE DISCHARGE
1. SILTATION Yes( ) No ( )
Areal Extent:
Depth:
2. VEGETATION GROWTH Yes ( ) No ( )
Areal Extent:
Type:
3. EROSION Yes( ) No ( )
Areal Extent:
Depth:
4. DISCHARGE STRUCTURE Yes ( ) No ( )
Functioning: Yes No
Condition:
FENCING
FENCING DAMAGE Yes ( ) No ( )
Description of damage:
PERIMETER ROADS
ROAD DAMAGE Yes ( ) No ( )
Description of damage:
SITE ACCESS
ACCESS RESTRICTIONS Yes ( ) No ( )
Description:
GENERAL
1. VANDALISM Yes ( ) No ( )
Description of damage:
2. CHANGED SITE CONDITION Yes ( ) No ( )
Description:
3. LAND USE CHANGE Yes ( ) No ( )
Description:
INTERVIEWS
1. INTERVIEW ON-SITE WORKERS Yes ( ) No ( )
Problems:
Suggestions:
Attach report:
2. INTERVIEW NEIGHBORS Yes ( ) No ( )
Droblems:
Suggestions:
Attach report:
3. INTERVIEW LOCAL OFFICIALS Yes ( ) No ( )
Droblems:
Suggestions:
Attach report:
DRAFT - DO NOT CITE OR QUOTE
9-7
-------
Table 9-2. Example of post-closure monitoring form used by U.S. Army Corps of
Engineers for CERCLA sites (cont).
REVIEW DOCUMENTS
1.GROUNDWATER MONITORING RECORDS
Abnormalities: Yes ( ) No ( )
2. GAS GENERATION RECORDS
Abnormalities: Yes ( ) No ( )
3. SETTLEMENT MONUMENT RECORDS
Abnormalities: Yes ( ) No ( )
4. OPERATION AND MAINTENANCE PLAN
Plan in Place? Yes ( ) No ( )
Plan is Being Followed? Yes ( ) No ( )
Plan is Adequate? Yes ( ) No ( )
Optimization is Being Considered? Yes ( ) No ( )
Systems with Optimization Potential? Yes ( ) No( )
9.2 Cover System Maintenance
9.2.1 Overview
There are a number of routine activities that should be conducted as part of a long-term cover
system maintenance program. These activities can generally be divided into the following major
categories:
• vegetation-related activities;
• erosion-related activities;
• subsidence-related activities;
• other surface layer performance related activities;
• drainage layer related maintenance;
• surface-water related activities; and
• monitoring system-related activities.
These maintenance categories, which are discussed in more detail below, are not all inclusive for
a facility. For example, site access control must also be maintained. In addition, for facilities
with gas control systems, there may be certain maintenance activities required under the CAA.
Further, there are likely other site-specific categories that need to be considered for waste
containment and remediation sites put to beneficial use.
DRAFT - DO NOT CITE OR QUOTE
-------
9.2.2 Vegetation-Related Maintenance
Cover system vegetation maintenance may include periodic irrigation and fertilization, as least
until vegetation is established, reseeding or replanting areas where vegetation has failed, cutting
young trees before they get too large and their roots disturb the cover system components, and
mowing. In virtually all cases, some degree of maintenance is necessary until the cover system
reaches a state of equilibrium with its inherent environment. Maintenance of cover system
vegetation is especially important for alternative cover systems that rely primarily on ET to limit
percolation.
As discussed in Section 2.2.3, grasses on cover systems located in humid or temperate climates
are usually mowed periodically to discourage the growth of deep-rooted plants, such as trees and
certain shrubs. Deep-rooted plants are usually undesirable because their root systems could plug
the drainage layer or penetrate and increase the hydraulic conductivity of the hydraulic barrier, if
the barrier consists of only a CCL or GCL without an overlying GM. Trees can also create
problems if they are blown over, uprooting large masses of soil and leaving a crater in the
surface. Many shrub species are shallow-rooted, do not require trimming/cutting, and are
sufficiently dense in their ground surface covering so as to prevent larger (deep-rooted) trees and
bushes from germinating. Mowing on a regular basis is expensive, thus its avoidance by proper
selection of shrub vegetation is an important design consideration.
9.2.3 Erosion-Related Maintenance
Cover system erosion, primarily by water, has been a problem for a number of cover systems, as
discussed in Section 2.2.5.1. It is important that significantly eroded areas be repaired in a
timely manner after they are observed to prevent progressive erosion and damage to cover
system components. Furthermore, it is easier to repair erosion rills prior to their development
into larger erosion gullies. As discussed in Section 2.2.5.2, rills can be removed by tilling the
soil surface. Gullies, on the other hand, generally cannot be repaired this way. Instead they
should be cut out and backfilled with soil that is blended into the adjacent soil.
9.2.4 Subsidence-Related Maintenance
As cover system settlement occurs, the surface grades of the cover system often decrease. If the
grades decrease substantially (and more than considered for design), the flow of water within
any cover system internal drainage layer and/or the flow of stormwater runoff may be impeded.
Regrading of a cover system is difficult not only from soil availability and placement
perspectives, but also from complications arising from pipes, piers, and other appurtenances
extending through the cover system. For example, a MSW landfill with an active gas extraction
system and leachate recirculation system may have numerous wells penetrating its cover system
and surface piping extending across the cover system, thereby requiring relatively small
construction equipment for maintenance regrading. Production rates with small equipment are
low. Obviously, the surface vegetation must be replaced after maintenance grading, and, in the
interval before vegetation is established, a temporary erosion control material may be necessary.
DRAFT - DO NOT CITE OR QUOTE
9-9
-------
The cover system may also exhibit localized differential settlements that cause ponding of water
and breaks in cover system piping. The existence of such depressions may lead to localized
areas with increased rates of percolation through the cover system. Whenever differential
settlement is visually observable, maintenance is necessary. If the cover system drainage layer,
hydraulic barrier, or finer-soil-to-coarser-soil interface, in the case of a capillary barrier, has also
subsided, the cover system will need to be reconstructed to bring the surface of these layers to
grade. For a capillary barrier, this repair must be carefully constructed, as described in Section
3.6.1, to reduce the potential for preferential pathways for infiltrating water. Besides causing
localized increases in percolation, cover system depressions also generate tensile strains in the
cover system components. As discussed in Section 2.5.2.5, tensile strains can cause barrier
materials to fail if the strains are excessive. Depending on the shape of the depression, and the
resulting tensile strains, a barrier material may need to be replaced in the depressed area. In
other words, bringing the surface of a CCL to grade in a depressed area will not be sufficient if
the CCL has failed due to excessive tensile strains. Instead, the barrier would have to be
repaired in some manner (e.g., by reconstructing the CCL or by bringing the CCL to grade and
placing a GM over the repaired area).
In addition to the above, subsidence-related maintenance may include adjusting the boots around
penetrations of the cover system barrier as the cover system settles.
9.2.5 Other Surface Layer Related Maintenance
To minimize percolation through the cover system, the integrity of the surface layer should be
maintained. Significant cracks or holes in the surface layer should be repaired, especially for
cover systems with ET or capillary barriers. The cracks may be caused by wet-dry cycles or may
be an indication of slope instability. Holes may be caused by burrowing animals.
9.2.6 Drainage Layer Related Maintenance
Drainage layer maintenance generally consists of clearing outlets of any obstacles, such as
debris, sediment or ice.
9.2.7 Maintenance of Surface-Water Management System
Maintenance of surface-water (i.e., stormwater) management systems is often required after
significant storm events. Excess sediment or other obstacles in drainage channels should be
removed, and damaged channel linings should be repaired. In areas where erosion has undercut
drainage channels (see Figure 7-19), the channels should be reconstructed. It is important that
these undercut areas are not just backfilled with soil if they are gully-like. As discussed above in
Section 9.2.3, gullies have to be cut out and reconstructed. Otherwise it is easier for the gully to
reform along the same flow path.
Drainage downchutes, outlets, energy dissipaters, and other areas where cover system
stormwater flows concentrate or substantially change energy state often require regular
maintenance and repair. These types of structures deserve careful attention during post-closure
DRAFT - DO NOT CITE OR QUOTE
9-10
-------
monitoring and need to be maintained in good operating condition. Gross et al. (2002) provide
several examples of damage to these types of structures resulting from stormwater flows.
9.2.8 Maintenance of Cover Monitoring System
Maintenance of the cover system monitoring system may include period re-calibration of
monitoring devices, replacement of batteries in data acquisition systems, and replacement of
damaged or non-functioning monitoring system components.
9.3 Site End Use
9.3.1 Overview
Increasingly, beneficial post-closure land uses are being considered in the design of cover
systems for waste containment facility closures and remediation sites. As of February 2001,
more than 190 cleaned up CERCLA sites have been returned to productive use (EPA, 200Ib).
EPA's Superfund Redevelopment Initiative reflects the Agency's belief that contaminated sites
should be cleaned up in a manner that is protective for reasonably anticipated future land use
(EPA, 1999a; EPA, 200 la). EPA does not favor one type of reuse over another, as land use is a
local decision. Further, the Agency believes that reuse should help to ensure proper maintenance
of the remedy (or cover system for waste containment sites) while providing tangible benefits to
key stakeholders, especially the surrounding community. The possible benefits of reuse include
(EPA, 1999a):
• "Positive economic impacts for communities living around the site including new
employment opportunities, increased property values, and catalysts for additional
redevelopment activities;
• Stakeholder acceptance of the municipal landfill presumptive remedy because of
potential time and cost savings, and increased involvement in the restoration and
redevelopment process;
• Enhanced day-to-day attention, potentially resulting in improved maintenance of remedy
integrity and institutional controls; and
• Improved aesthetic quality of the area through discouragement of illegal waste disposal
or trespassing on restricted portions of the site, as well as increased upkeep of the site by
future site occupants."
For CERCLA sites, EPA must balance this preference for future land use with other technical
and legal provisions, including ARARs. Only if the remedy is anticipated to achieve cleanup
levels that allow the site to be available for the reasonable anticipated future land use, will EPA
support that reuse.
The reuse selected for a given site is a function of a number of factors, including the
stakeholders, site features, environmental considerations, site ownership, land use considerations
and environmental regulations, community input, and public initiatives. These factors are
DRAFT - DO NOT CITE OR QUOTE
9-11
-------
discussed in EPA (200la). The three major categories of site end use that have been employed
at waste containment facilities and remediation sites are: (i) ecological enhancement; (ii)
recreational reuse; and (iii) industrial and commercial reuse (EPA, 1999a). Each of these
categories is discussed in more detail below, and case histories illustrating these categories are
presented. Additional detail is provided in EPA publications (available for download at the EPA
website http://www.epa.gov/superfund/programs/recycle/newdocs.htm) on the recreational reuse
(EPA, 200 Ib) and commercial reuse (EPA, 2002) of CERCLA sites. About half of the 190
CERCLA sites mentioned above that had been developed by February 2001 are being used for
industrial or commercial purposes (EPA, 2002).
Whatever the type of end use, there are site design issues, such as settlement, gas management,
and surface-water management, which are often common to many sites. In addition, some types
of sites and end uses may have more issues than others. For example, when developing a former
MSW landfill site as a retail shopping complex, there is extra concern about foundation
settlement and gas migration to enclosed structures. If the site were developed as wildlife
habitat, settlement and gas migration would likely not be as much a concern.
The selected end use can have a significant impact on cover system design. For example, if a
site is to be used for a golf course or other facility with a vegetated surface layer that requires
irrigation, the cover system may require an internal drainage layer and a barrier that includes a
GM to control percolation through the cover system. It is important that the site end use be
considered during the design phase of the cover system so that any special features needed to
support the post-closure use can be incorporated into the cover system at that time. It can be
significantly more expensive to retrofit a constructed cover system to support a specific site end
use than to design the cover system to support the specific end use from the start. These end-use
designs will have their own monitoring and maintenance requirements. Personnel maintaining
the end-use facility should be aware of the maintenance requirements related to the prior
disposition of the facility (i.e., waste containment facility or remediation site).
9.3.2 Ecological Reuse
Closed waste containment and remediation sites located in ecologically significant areas have
been used as wildlife restoration areas or wetlands. Besides providing a nurturing environmental
for plants and wildlife, wetlands filter sediments and contaminants from surface water and can
absorb floodwaters, which reduces the flooding potential for lowlands.
9.3.3 Recreational Reuse
Closed MSW landfills are a natural fit for reuse as recreation areas because they typically have a
large surface area, and the cover system can generally be contoured to meet the specifications for
recreational facilities, such as ball fields or golf courses (EPA, 2001b). Recreational reuse has
included trails for hiking, mountain biking, or horseback riding, camping facilities, picnic areas,
parks, playgrounds, sledding areas, playgrounds, ball fields, and golf courses. In many cases, a
site that will be developed for recreational purposes will support more than one type of
recreational activity. For example, a site developed as a general use park may also accommodate
DRAFT - DO NOT CITE OR QUOTE
9-12
-------
sports fields, playgrounds, trails, or other recreational features. In other cases, recreation may be
secondary to a primary use, such as a commercial development. Detailed information on the
development of recreation facilities over waste containment facilities and remediation sites is
presented in EPA (200Ib) and is not repeated herein.
9.3.4 Industrial and Commercial Reuse
The beneficial use of closed sites is particularly attractive in areas where developable real estate
is limited and expensive. In major urban areas, closed waste containment and remediation sites
are increasingly viewed as offering potential for traditional urban developments, such as office
parks and retail centers. In such settings, these facilities may not be suitable for ecological or
recreational use. Industrial and commercial reuse has included parking lots, restaurants, retail
shopping stores or complexes, office buildings, intermodal transportation facilities, port cargo
handling facilities, and airports.
One impediment to the design of structures over closed waste containment facilities or
remediation sites is that the underlying materials (waste or contaminated materials) may have
much different properties than soil. The foundations for these structures should be carefully
designed to be protective of the cover system and prevent structural damage. If the waste or
contaminated material is anticipated to experience large settlements (e.g., as is typical for
MSW), the use of shallow building foundations (e.g., spread footings, reinforced concrete mats,
grid foundations with column footings tied together with a system of grade beams and usually an
integrated concrete floor) is generally limited to small lightly loaded structures that can tolerate
some differential settlements (Dunn, 1995). These shallow foundations are typically located
above the cover system barrier layer and contain more reinforcing steel than is required for
foundations on conventional sites. Structures on shallow foundations can also be designed to
accommodate differential settlements by using tilt-up wall construction, where both the wall
sections and the footings are broken up into discrete sections with control and leveling joints
between them, by casting the slab in separate sections connected by cable linkages, or by other
means (EPA, 2002).
If settlements are anticipated to be too high, site improvement techniques can be considered.
Dunn (1995) offers these techniques for reducing the total settlement of structures constructed
over MSW landfills:
• allowing the MSW to reach an acceptable level of decomposition, either by delaying
construction or enhancing decomposition ... ;
• supplemental compaction of the MSW, which is usually limited to relatively shallow
MSW depths of no more than two or three meters;
• surcharging, with settlement monitoring;
• dynamic compaction; and
• grouting or fly-ash injection.
DRAFT - DO NOT CITE OR QUOTE
9-13
-------
If these techniques are unfeasible, deep foundations can be considered.
Heavier structures over waste materials may need to be supported on deep foundations, which
are typically piles driven into competent supporting materials below the waste, though drilled
piers are also sometimes used. Deep foundations may not be appropriate for sites with a liner
system, with wastes that are difficult to drive or drill through, or that have an uncontaminated
aquifer that could be impacted by the foundation construction (EPA, 2002). Where deep
foundations penetrate the cover system, the penetrations need to be carefully designed to control
infiltration and gas emissions. In some cases, structures on pile or pier foundations may settle
less than the surrounding area, and gaps may form between the structure and adjacent features
(e.g., roads, parking lots, etc.), potentially damaging structure entryways and utilities. Periodic
maintenance of these structures may include site regrading, repair of entrances, and adjustment
of utilities.
Shallow and deep foundations on waste containment or remediation sites are designed using
standard geotechnical methods with consideration of settlement, bearing capacity of shallow
foundations, capacity of deep foundations, and downdrag due to waste settlement. In addition to
these geotechnical considerations, environmental factors, and especially gas migration, must be
considered. Gas migration to enclosed structures is especially a concern with site reuse. Sites
that are expected to produce significant amounts of gas may not be good candidates for industrial
or commercial uses, unless the gas is well controlled. For this case, there are generally two
systems for gas control: (i) a gas management system that is usually incorporated into the
containment system; and (ii) a gas protection system for the structure that is usually independent
of the gas management system. Gas protection techniques used for industrial and commercial
facilities include (EPA, 2002):
• "Construct floor slabs with convex bottoms to prevent methane from pooling below the
structure.
• Place an impermeable (gas resistant) geomembrane or other hydraulic/gas barrier under
the structure or within the building's floors. This is especially important for sites likely
to experience settlement that may disrupt the cover.
• Engineer an air space below a structure to allow for gas detection and venting, as well
as to facilitate inspection and maintenance of the cover.
• Place gas detectors in closed structures to warn of potential gas buildup.
• Install vent fans to remove methane buildup from the structure.
• Ensure that the design of utilities does not allow for gas migration along utility conduits.
One approach is to attach utility service entrances to the outside wall of the structure so
they do not penetrate the floor slab, which may create a pathway for gas entry. "
Additional detail on the development of commercial facilities over waste containment facilities
and remediation sites is presented in EPA (2002) and is not repeated herein. Most of this
information is also applicable to the development of industrial facilities over these sites.
DRAFT - DO NOT CITE OR QUOTE
9-14
-------
9.3.5 Case Histories
Several published case histories of different site end uses for different types of facilities are
presented below. Additional case histories are presented in several EPA publications (1999a,
200Ib, 2002), which can be downloaded from the Agency's website at
http://www.epa.gov/superfund/programs/recycle/product.htm. The website also include
individual case histories of the reuse of some CERCLA sites.
Bowers Landfill
As described by EPA (1999a, 1999b), the 5-ha Bowers Landfill site was located in a former rock
quarry within the Scioto River floodplain in central Ohio. Municipal, chemical, and industrial
wastes were disposed in the landfill. Until the remedy was constructed the site was flooded an
average of 29 days/yr, and contaminants from the site were carried to groundwater and the river.
The remedy included removing surface debris and sediments, constructing a cover system that
included a CCL barrier and gas collection system over the landfill, and creating 3-ha of wetlands
between the landfill and the river. The wetlands not only provide a protective buffer between the
landfill and river, but also provide habitat for numerous species of plants, birds, and other
wildlife.
Figure 9-1. Constructed Wetlands at Bowers Landfill Site.
DRAFT - DO NOT CITE OR QUOTE
9-15
-------
Three Landfills in Florida
Mackey (1996) presents case studies of different end uses that were implemented at three closed
landfills in Florida. The first site, the Key Largo Landfill Facility, was developed as a nature
preserve. This 6.0-ha facility is surrounded on three sides by the Florida Crocodile Refuge,
which is maintained by the U.S. Fish and Wildlife Service (USFWS) and provides habitat for
several endangered species. The cover system design consists of, from top to bottom:
• 0.15-m thick vegetated topsoil surface layer;
• 0.30-m thick limerock protection layer;
• GC drainage layer;
• 1.0-mm thick textured HDPE/VLDPE/HDPE GM; and
• 0.15-m thick compacted limerock foundation layer.
To enhance the value of the facility as a wildlife preserve, the cover system was vegetated with
native plant species and feral cats and certain exotic plants were removed.
The second site, the 13.2-ha Sanlando Landfill Facility, was developed into a softball complex.
During the selection of an end-use for the site, it was anticipated that a softball complex would
get significant use because it would be located adjacent to a park already used by the community
and there had been a large growth in population in the vicinity of the facility. Due to the
numerous cover system penetrations that would be required to install poles for fencing and
lights, piers for buildings constructed over the landfill, and utility conduits, the design engineers
decided to use a CCL hydraulic barrier rather than a GM barrier over the majority of the facility.
However, beneath buildings, a 0.5-mm thick PVC GM barrier was installed to reduce the
potential for gas migration into the structures.
The third site, the Dyer Boulevard Landfill Facility, occupies 260 ha and includes three large
disposal areas, one area containing MSW, one area containing construction and demolition waste
(C&DW), and the remaining area containing mixed waste, waste and C&DW. This facility was
developed into a multi-faceted sports and recreation complex that includes basketball courts,
soccer fields, tennis courts, an observatory mound, picnic areas, canoe rentals, and multi-purpose
trails for pedestrians, joggers, bikers, and horses. A specific design feature was incorporated into
the cover system over the C&DW area. The end use of this area was a mountain bike challenge
course. However, there was concern that the mountain bikes would cause rutting and erosion of
the cover soils. To monitor the effect of the activity on the cover system and limit any
significant impact, a GT reinforcement layer was placed beneath the mountain bike trails. The
purpose of the GT is twofold: (i) to reduce rutting; and (ii) to alert maintenance personal that
significant rutting has occurred (and that the cover system must be repaired).
DRAFT - DO NOT CITE OR QUOTE
9-16
-------
Chisman Creek Superfund Site
As described by EPA (1999a, 1999c, 2001b), the 11-ha Chisman Creek Superfund site is located
in York County, Virginia near Chisman Creek, a tributary of Chesapeake Bay. From 1957 to
1974, more than 500,000 tonnes of fly ash from a coal-fired power plant was deposited into
abandoned sand and gravel pits on the site. The fly ash was not covered, and eventually resulted
in groundwater, surface-water, and soil contamination. The remedy included constructing a
cover system over the fly ash and installing a leachate collection and treatment system in the
oldest and deepest pit. Because fly ash has low compressibility and doesn't generate gas, fly ash
fills can be ideal sites for structures.
The site, plus some adjacent property, was developed into two sports parks, with two lighted
softball fields, four soccer fields, restrooms, vending facilities, equipment storage facilities, and
a parking area (Figure 9-2). The cover system was engineered to serve as a foundation for the
park facilities and graded to accommodate park structures. The cover system design consists of,
from top to bottom:
• 0.15-m thick vegetated topsoil surface/protection layer;
• 0.15-m thick sand drainage layer;
• 0.3-m thick CCL; and
• 0.3-m thick soil/ash mixture.
Figure 9-2. Softball Fields at Chisman Creek Superfund Site (from EPA, 1999c).
Precautions, such as placing underground utilities in oversized trenches filled with clean fill,
were taken to protect future workers from coming into contact with the fly ash.
DRAFT - DO NOT CITE OR QUOTE
9-17
-------
McColl Super fund Site
As described by Collins et al. (1998), the 8.8-ha McColl Superfund site is located in Fullerton,
California and includes 12 unlined pits of sludges and other wastes from production of high-
octane aviation fuel (Figure 9-3(a)). In the 1950s and 1960s, three pits were covered with diesel-
oil based drilling mud and six pits were covered with soil to control odor and gaseous emissions.
The site was placed on the NPL in 1982. The remedy for the site was designed around its end
use as part of the Los Coyotes Golf Course and wildlife sanctuary (Figures 9-3(b) and (c)). The
remedy includes a multi-component soil and geosynthetic cover system designed to control
infiltration and emissions of thiophene compounds, retaining walls to stabilize steep slopes
adjacent to the pits, and a soil-bentonite slurry wall. In areas that had been covered with soft
drilling muds, a lightweight geocell-reinforced cover system was used. Beneath the golf course,
the cover system was geogrid reinforced and included a cobble protection layer to minimize the
potential for human intrusion. For both conditions, the cover system included an HDPE
GM/GCL barrier underlain by a sand gas collection/foundation layer.
DRAFT - DO NOT CITE OR QUOTE
9-18
-------
(c)
Figure 9-3. McColl Superfund Site: (a) Before Closure; (b,c) After Development as a Golf
Course and Wildlife Sanctuary ((c) was downloaded from an EPA website at
http://www.epa. gov/superfund/programs/recycle/briefs/ca_brief.htm).
DRAFT - DO NOT CITE OR QUOTE
9-19
-------
Raymark Super fund Site
As discussed by EPA (1999a, 1999d, 2002), the 14-ha Raymark Superfund site is located in
Fairfield County, Connecticut and was operated from 1919 to 1989 as a manufacturing facility
for automotive parts and products. Waste generated during the assembly process was disposed
in on-site lagoons. As these lagoons reached capacity, they were dredged and the dredged
materials were used as fill for construction on over 70 local properties, including school playing
fields, recreational parks, and commercial and residential properties. The dredged materials
contained lead, asbestos, PCBs, dioxins, and 60 other hazardous substances, and subsequently
contaminated soil and groundwater. The remedy for the contaminated properties consisted of
relocating contaminated materials back to the Raymark Superfund site or constructing cover
systems over them. On the Raymark property, buildings and structures within a 6-ha area were
decontaminated and demolished, a groundwater pump-and-treat system was installed, and the
on-site and off-site contaminated soils were collected and contained with a cover system. The
cover system included GM/CCL hydraulic barrier and underlying sand gas collection layer.
Between 0.9 and 3 m of clean fill were placed over the hydraulic barrier to facilitate site
development and protect the barrier.
Figure 9-4. Conceptual Drawing of Future Shopping Center at Raymark Superfund
Site (EPA, 1999d).
The proposed end-use for the Raymark Superfund site is a 3-ha retail shopping center (Figure
9-4). Prior to construction of the cover system, the site was improved to enhance its
geotechnical properties. In-place soils and waste were stabilized using dynamic compaction or
surcharging, and peat deposits were dewatered using wick drains. A 0.2-ha platform foundation
for the shopping center has been constructed. The platform is supported by 277 30-m long piles
that penetrate the cover system.
DRAFT - DO NOT CITE OR QUOTE
9-20
-------
Denver Radium Super fund Site
As described by EPA (1999e, 2002), operable unit (OU) 9 of the Denver Radium Superfund site
is a 7-ha property located in Denver, Colorado that was first used for industrial activities in
1886, with the construction of a smelter. The site was subsequently used for other industrial
activities, including cyanide leaching, zinc milling, radium ore processing, minerals recovery,
manufacturing and servicing of batteries, oil reclamation, and brick manufacturing. As property
ownership, industrial activities, and land use changed, radioactive by-products were often left in
place, used as fill or foundation material, or otherwise mishandled. At the time the site remedy
was selected, the site soil was contaminated with radium-226, arsenic, zinc, and lead.
The remedy for OU 9 consisted of excavating radioactive materials found at the site and shipping
them to an offsite disposal facility and relocating 13,000 m3 of metals-contaminated soils to four
unlined containment cells covered with asphaltic concrete. The primary risks to human health
and the environment posed by the soils are related to the ingestion or inhalation of the metals.
Since the metals in the soils are only slightly soluble, percolation of water through the soils is
not likely to cause the metals to migrate. Thus, the cover systems for the four cells were
designed to minimize contact with the waste, rather than to minimize percolation. The remedy
was developed concurrently with the design of the site reuse. The site has been developed with
a large retail store and parking lot (Figure 9-5). The uncontaminated spaces between the four
containment cells were used for utility corridors, and the asphaltic concrete cover systems were
incorporated into the parking lot. The store, itself, was constructed on uncontaminated soil.
Figure 9-5. Part of the Denver Radium Superfund Site was Developed with a Retail
Store (EPA, 1999e).
DRAFT - DO NOT CITE OR QUOTE
9-21
-------
Appendix A
References by Chapter
Chapter 1
Bonaparte, R. (1995). "Long-Term Performance of Landfills," Geoenvironment 2000,
Geotechnical Special Publication No. 46, D.E. Daniel and Y.B. Acar (eds.), ASCE, New York,
NY, Vol. l,pp. 514-553.
Caldwell, J.A. and Reith, C.C. (1993). "Principles and Practice of Waste Encapsulation" Lewis
Publishers, Chelsea, MI, 414 p.
EPA (1989). "Final Covers on Hazardous Waste Landfills and Surface Impoundments"
Technical Guidance Document, EPA/530/SW-89/047, U.S. Environmental Protection Agency,
Office of Solid Waste and Emergency Response, Washington, D.C., 39 p.
EPA (199la). "Design and Construction of RCRA/CERCLA Final Covers," Seminar
Publication, EPA/625/4-91/025, U.S. Environmental Protection Agency, Office of Research and
Development, Washington, D.C., 145 p.
EPA (1991b). "Conducting Remedial Investigations/Feasibility Studies for CERCLA Municipal
Landfill Sites" EPA/540/P-91/001, U.S. Environmental Protection Agency, Office of Emergency
and Remedial Response, Washington, D.C., 171 p.
EPA (1991c). Appendix E to Final Rule, 40 CFR Parts 257 and 258, "Solid Waste Disposal
Facility Criteria; Final Rule," Federal Register, Vol. 56, No. 196, pp. 51057-51060.
EPA (1992a). 40 CFR Parts 260, 264, 265, 270, and 271, "Liners and Leak Detection Systems
for Hazardous Waste Land Disposal Units; Final Rule," Federal Register, Vol. 57, No. 19, pp.
3462-3497.
EPA (1992b). "Solid Waste Facility Disposal Criteria; Final Rule Corrections," Federal
Register, Vol. 57, No. 124, pp. 28626-28629.
EPA (1993). "Presumptive Remedy for CERCLA Municipal Landfill Sites," OSWER Directive
No. 9355.0-49FS, EPA/540/F-93/035, U.S. Environmental Protection Agency, Office of Solid
Waste and Emergency Response, Washington, D.C., 14 p.
EPA (1995). " Landfill Bioreactor Design and Operation" Seminar Publication, EPA/600/R-
95/146, U.S. Environmental Protection Agency, National Risk Management Research
Laboratory, Cincinnati, OH, 230 p.
EPA (1997a). "Clean Up the Nation's Waste Sites: Markets and Technology Trends," 1996
Edition, EPA/542/R-96/005, U.S. Environmental Protection Agency, Office of Solid Waste and
Emergency Response, Washington, D.C.
DRAFT - DO NOT CITE OR QUOTE
A-l
-------
EPA (1999a). "Use of Monitored Natural Attenuation at Superfund, RCRA Corrective Action,
and Underground Storage Tank Sites" OSWER Directive 9200.4-17P, U.S. Environmental
Protection Agency, Office of Solid Waste and Emergency Response, Washington, D.C., 32 p.
EPA (1999b). "Field Applications of In Situ Remediation Technologies: Permeable Reactive
Barriers" EPA/542/R-99/002, U.S. Environmental Protection Agency, Office of Solid Waste
and Emergency Response, Washington, D.C., 114 p.
EPA (2000). 40 CFR Part 63, "National Emission Standards for Hazardous Air Pollutants:
Municipal Solid Waste Landfills; Proposed Rule," Federal Register, Vol. 65, No. 216, pp.
66672-66686.
EPA (2000a). "Revised Alternative Cap Design Guidance for Unlined, Hazardous Waste
Landfills in EPA Region 7," Internal EPA Region 1 Memorandum, 05 February 2001.
Giroud, J.P. and Bonaparte, R. (1989). "Leakage Through Liners Constructed with
Geomembranes, Part II: Composite Liners," Geotextiles and Geomembranes, Vol. 8, No. 2, pp.
71-111.
Gleason, M.H., Houlihan, M.F., and Giroud, J.P. (1998). "An Exposed Geomembrane Cover
System for a Landfill," Proceedings, Sixth International Conference on Geosynthetics, Industrial
Fabrics Association International, St. Paul, MN, Vol. 1, pp. 211-218.
Gleason, M.H., Germain, A.M., Vasuki, N.C., Giroud, J.P. (1999). "Design and Construction of
an Exposed Geomembrane Cover System for a Solid Waste Landfill", Proceedings, Sardinia
'99, Seventh International Waste Management and Landfill Symposium, Caligliari, Italy, pp.
335-342.
Gleason, M.H., Houlihan, M.F., and Palutis, J.R. (2001). "Exposed Geomembrane Cover
Systems: Technology Summary", Geosynthetics '01 Conference Proceedings, Industrial Fabrics
Association International, St. Paul, MN, pp. 905-918.
Gross, B.A., Bonaparte, R., and Giroud, J.P. (2002). "Waste Containment Systems: Problems
and Lessons Learned," Appendix F in Assessment and Recommendations for Optimal
Performance of Waste Containment Systems, R. Bonaparte, D.E. Daniel, and R.M. Koerner, U.S.
Environmental Protection Agency, National Risk Management Research Laboratory, Cincinnati,
OH.
Hakonson, T.E., Lane, L.G., and Springer, E.P. (1992). "Biotic and Abiotic Processes," Chapter
4 in Deserts as Dumps?, C.C. Reith and B.M. Thomson (eds.), University of New Mexico Press.
Harper, J.L. (1987). "The Heuristic Value of Ecological Restoration," Restoration Ecology: A
Synthetic Approach to Ecological Research, W.R. Jordan III, M.E. Gilpin, and J.B. Aber (eds.),
Cambridge University Press, New York, NY, pp. 35-45.
Hauser, V.L., Gimon, D.M., and Jackson, D.R. (2000). "Golf Courses on Air Force Landfills,"
Air Force Center for Environmental Excellence, Technology Transfer Division, Brooks AFB,
TX, 14 p.
DRAFT - DO NOT CITE OR QUOTE
A-2
-------
Hillel, D. (1998). "Environmental Soil Physics" Academic Press, San Diego, CA, 771 pp.
Hsuan, Y.G. and Koerner, R.M. (1998). "Antioxidant Depletion Lifetime in High Density
Polyethylene Geomembranes," Journal of Geotechnical and Geoenvironmental Engineering,
ASCE, Vol. 124, No. 6, pp. 532-541.
Hsuan, Y.G. and Koerner, R.M. (2002). "Antioxidant Depletion Time in High Density
Polyethylene Geomembranes," Appendix B in "Assessment and Recommendations for Optimal
Performance of Waste Containment Systems" R. Bonaparte, D.E. Daniel, and R.M. Koerner,
U.S. Environmental Protection Agency, National Risk Management Research Laboratory,
Cincinnati, OH.
Koerner, R.M. and Daniel, D.E. (1997). "Final Covers for Solid Waste Landfills and Abandoned
Dumps" ASCE Press, Reston, VA, 256 p.
Link, S.O., Waugh, W.J., and Downs, J.L. (1994). "The Role of Plants in Isolation Barrier
Systems," In-situ Remediation: Scientific Basis for Current and Future Technologies., G.W. Gee
and N.R. Wing (eds.), Batelle Press, Richland, WA, pp. 561-592.
Mitchell, J.K. and Jaber, M. (1990). "Factors Controlling the Long-Term Properties of Clay
Liners," Waste Containment Systems: Construction, Regulation, and Performance., Geotechnical
Special Publication No. 26, R. Bonaparte (ed.), ASCE, New York, NY, pp. 84-105.
Othman, M.A., Bonaparte, R., Gross, B.A., and Warren, D. (2002). "Evaluation of Liquids
Management Data for Double-Lined Landfills," Appendix E in Assessment and
Recommendations for Optimal Performance of Waste Containment Systems., R. Bonaparte, D.E.
Daniel, and R.M. Koerner, U.S. Environmental Protection Agency, National Risk Management
Research Laboratory, Cincinnati, OH.
Tchobanoglous, G., Theisen, H., and Virgil, S.A. (1993). "IntegratedSolid Waste
Management" McGraw-Hill, New York, 992 p.
Thorneloe, S., Roquetta, A., Pacey, I, and Bottero, C. (2000). "Database of Landfill-Gas-to-
Energy Projects in the United States," MSWManagement, Vol. 10, No. 2, pp. 29-37.
Waugh, WJ. (1994). "Paleoclimatic Data Application: Long-Term Performance of Uranium
Mill Tailings Repositories," Workshop Proceedings: Climate Change in the Four Corners and
Adjacent Regions, Grand Junction, CO, Sept 12-14.
Whittaker, R.H. (1975). "Communities and Ecosystems" 2nd Edition, MacMillan Publishing
Co., Inc., New York, NY.
Chapter 2
Abt, S.R., Khattak, M.S., Nelson, J.D., Ruff, J.F., Shaikh, A., Wittier, R.J., Lee, D.W., and
Hinkle, N.E. (1987). "Development of Riprap Design Criteria by Riprap Testing in Flumes:
Phase I" NUREG/CR-4651, U.S. Nuclear Regulatory Commission, Washington, DC, 111 p.
DRAFT - DO NOT CITE OR QUOTE
A-3
-------
Abt, S.R., Wittier, R.J., Ruff, J.F., LaGrone, D.L., Khattak, M.S., Nelson, J.D., Hinkle, N.E., and
Lee, D.W. (1988). "Development of Riprap Design Criteria by Riprap Testing in Flumes: Phase
IIB Follow-up Investigations" NUREG/CR-4651-V2, U.S. Nuclear Regulatory Commission,
Washington, DC, 84 p.
Albrecht, B.A. and Benson, C.H. (2001). "Effect of Desiccation on Compacted Natural Clays,"
Journal of Geotechnical and GeoenvironmentalEngineering, Vol. 127, No. 1, pp. 67-75.
Aldrich, H.P. and Paynter, H.M. (1953). "Analytical Studies of Freezing and Thawing of Soils,"
Frost Investigations, Fiscal Year 1953, First Interim Report., Technical Report 42, U.S. Army
Corps of Engineers, Artie Construction and Frost Effects Laboratory, New England Division,
Boston, MA.
Alston, C., Daniel, D.E., and Devroy, DJ. (1997). "Design and Construction of Sand-Bentonite
Liner for Effluent Treatment Lagoon, Marathon, Ontario," Canadian Geotechnical Journal, Vol.
34, No. 6, pp. 841-862.
Barfield, B.J., Warner, R.C., and Haan, C.T. (1983). "AppliedHydrology andSedimentology for
Disturbed Areas, " Oklahoma Technical Press.
Benson, C.H., Daniel, D.E., and Boutwell, G.P. (1999). "Field Performance of Compacted Clay
Liners," Journal of Geotechnical and GeoenvironmentalEngineering., Vol. 125, No. 5, pp. 390-
403.
Boardman, B.T., and Daniel, D.E. (1996). "Hydraulic Conductivity of Desiccated Geosynthetic
Clay Liners," Journal of Geotechnical Engineering., Vol. 122, No. 3, pp. 204-208.
Bonaparte, R., Othman, M.A., Rad, N.S., Swan, R.H., and Vander Linde, D.L. (1996).
"Evaluation of Various Aspects of GCL Performance," Appendix F in Report of 1995 Workshop
on Geosynthetic Clay Liners, D.E. Daniel and H.B. Scranton (eds.), EPA/600/R-96/149, U.S.
Environmental Protection Agency, National Risk Management Research Laboratory, Cincinnati,
OH, pp. F1-F34.
Bowders, J.J., Daniel, D.E., Wellington, J., and Houssidas, V. (1997). "Managing Desiccation
Cracking in Compacted Clay Liners Beneath Geomembranes," Geosynthetics '97 Conference
Proceedings, Industrial Fabrics Association International, St. Paul, MN, Vol. 1, pp. 527-540.
Boynton, S.S. and Daniel, D.E. (1985). "Hydraulic Conductivity Tests on Compacted Clays,"
Journal of Geotechnical Engineering, Vol. Ill, No. 4, pp. 465-478.
Caldwell, J.A. and Reith, C.C. (1993). "Principles and Practice of Waste Encapsulation" Lewis
Publishers, Chelsea, MI, 414 p.
Cassel, O.K. and Nielsen, D.R. (1986). "Field Capacity and Available Water Capacity, Methods
of Soil Analysis, Part 1," Physical andMineralogicalMethods, Agronomy Monograph No. 9, 2
Edition, Soil Society of America, Madison, WI.
Chow, V.T. (1959). "Open Channel Hydraulics," McGraw-Hill, New York, NY.
DRAFT - DO NOT CITE OR QUOTE
A-4
nd
-------
Cline, J.F. (1979). "Biobarriers Used in Shallow-Burial Ground Stabilization" PNL-2918,
Pacific Northwest Laboratory, Richland, WA.
Cline, J.F., Cataldo, D.A., Burton, F.G., and Skiens, W.E. (1982). "Biobarriers Used in Shallow
Burial Ground Stabilization," Nuclear Technology., Vol. 58, pp. 150-153.
Comer, A.I., Sculli, M.L., and Hsuan, Y.G. (1995). "Effects of Freeze-Thaw Cycling on
Geomembrane Sheets and Their Seams," Geosynthetics '95 Conference Proceedings, Industrial
Fabrics Association International, St. Paul, MN, pp. 853-866.
Corser, P. and Cranston, M. (1991). "Observations on Long-Term Performance of Composite
Clay Liners and Covers," Proceedings, Geosynthetic Design and Performance, Vancouver
Geotechnical Society, Vancouver, BC.
Corser, P., Pellicer, J., and Cranston, M. (1992). "Observation on Long-Term Performance of
Composite Clay Liners and Covers," Geotechnical Fabrics Report, Vol. 10, No. 8, Industrial
Fabrics Association International, St. Paul, MN, pp. 6-16.
Daniel, D.E. and Benson, C.H. (1990). "Water Content-Density Criteria for Compacted Clay
Liners," Journal of Geotechnical Engineering, Vol. 116, No. 12, pp. 1811-1830.
Daniel, D.E. and Boardman, B.T. (1993). "Report of Workshop on Geosynthetic Clay Liners"
EPA 600/R-93/171, U.S. Environmental Protection Agency, Risk Reduction Engineering
Laboratory, Cincinnati, OH, 106 p.
Daniel, D.E. (2002). "Cincinnati Geosynthetic Clay Liner Test Site," Appendix D in Assessment
and Recommendations for Optimal Performance of Waste Containment Systems, R. Bonaparte,
D.E. Daniel, and R.M. Koerner, U.S. Environmental Protection Agency, National Risk
Management Research Laboratory, Cincinnati, OH (in press).
Daniel, D.E. and Estornell, P.M. (1991). "Compilation of Information on Alternative Barriers
for Liner and Cover Systems," EPA/600S/2-91/002, U.S. Environmental Protection Agency,
Risk Reduction Engineering Laboratory, Cincinnati, OH.
Daniel, D.E. and Koerner, R.M. (1993). "Technical Guidance Document: Quality Assurance
and Quality Control for Waste Containment Facilities," EPA/600/R-93/182, U.S. Environmental
Protection Agency, Risk Reduction Engineering Laboratory, Cincinnati, OH, 305 p.
Daniel, D.E. and Koerner, R.M. (1995). "Waste Containment Facilities: Guidance for
Construction Quality Assurance and Quality Control of Liner and Cover Systems," ASCE Press,
New York, NY, 354 p.
Daniel, D.E. and Scranton, H.B. (1996). "Report of 1995 Workshop on Geosynthetic Clay
Liners," EPA/600/R-96/149, U.S. Environmental Protection Agency, National Risk Management
Research Laboratory, Cincinnati, OH, 96 p.
Daniel, D.E. and Wu, Y.K. (1993). "Compacted Clay Liners and Covers for Arid Sites," Journal
of Geotechnical Engineering, Vol. 119, No. 2, pp. 223-227.
DRAFT - DO NOT CITE OR QUOTE
A-5
-------
Daniel, D.E., Shan, H.-Y., and Anderson, J.D. (1993). "Effects of Partial Wetting on the
Performance of the Bentonite Component of a Geosynthetic Clay Liner," Geosynthetics '93
Conference Proceedings, Industrial Fabrics Association International, St. Paul, MN, Vol. 3, pp.
1483-1496.
DeGaetano, A.T., Wilks, D.S., and McKay, M. (1997). "Extreme-Value Statistics for Frost
Penetration Depths in Northeastern Unites States" Journal of Geotechnical and
GeoenvironmentalEngineering, Vol. 123, No. 9, pp. 828-835.
DePoorter, G.L. (1982). "Shallow Land Burial Technology Development" presentation to the
Low-Level Waste Management Program Review Committee, Los Alamos National Laboratory,
Los Alamos, NM.
DOE (1989). "Technical Approach Document, Revision II," Uranium Mill Tailings Remedial
Action Project UMTRA-DOE/AL 050425.0002, U.S. Department of Energy Albuquerque, NM,
290 p.
EPA (1985). "Covers for Uncontrolled Hazardous Waste Sites," U.S. Environmental Protection
Agency, Hazardous Waste Engineering Research Laboratory, EPA/540/2-85/002, 561 p.
EPA (1991). "Design and Construction of RCRA/CERCLA Final Covers" Seminar Publication,
EPA/625/4-91/025, U.S. Environmental Protection Agency, Office of Research and
Development, Washington, D.C., 145 p.
EPA (2002). "Assessment and Re commendations for Optimal Performance of Waste
Containment Systems" EPA/600/R-02/099, U.S. Environmental Protection Agency, National
Risk Management Research Laboratory, Cincinnati, OH.
Erickson, A.E., Chamberlain, E.J., and C.H. Benson (1994). "Effects of Frost Action on Covers
and Liners Constructed in Cold Environments," Proceedings, 17th Madison Waste Conference,
University of Wisconsin, Madison, WI, pp. 198-220.
Foose, G.J., Benson, C.H., and Edil, T.B. (2001). "Predicting Leakage Through Composite
Landfill Liners," Journal of Geotechnical and Geoenvironmental Engineering, Vol. 127, No. 6,
pp. 510-520.
Fox, PJ. (1998). "Research on Geosynthetic Clay Liners at Purdue University," Geotechnical
News, March, pp. 35-40.
Fryrear, D.W., Saleh, A., Bilbro, J.D., Schomberg, H.M., Stout, I.E., andZobeck, T.M. (1998).
"Revised Wind Erosion Equation (RWEQ)," Wind Erosion and Water Conservation Research
Unit, U.S. Department of Agriculture, Agricultural Research Service, Southern Plains Area
Cropping Systems Research Laboratory, Technical Bulletin No. 1.
Garrels, R.M. (1951). "A Textbook of Geology", Harper Brothers, New York, NY, Slip.
Gee, G.W., Payer, M.J., Rockhold, M.L., and Campbell, M.D. (1992). "Variations in Recharge
at the Hanford Site," Northwest Science, Vol. 66, pp. 237-250.
DRAFT - DO NOT CITE OR QUOTE
A-6
-------
Gerber, D.H. (1984). "Floating Reservoir Cover Designs", Proceedings, International
Conference on Geomembranes, Industrial Fabrics Association International, St. Paul, MN, Vol.
1, pp. 79-84.
Gilbert, P.A. and Murphy, W.L. (1987). "Prediction/Mitigation of Subsidence Damage to
Hazardous Waste Landfill Covers" EPA/600/12/87-025, U.S. Environmental Protection
Agency, Hazardous Waste Engineering Research Laboratory, 81 p.
Giroud, J.P. (1994). "Quantification of Geosynthetics Behavior," Proceeding of the Fifth
International Conference on Geotextiles, Geomembranes, and Related Products, Singapore, Vol.
4, pp. 1249-1273.
Giroud, J.P. (1997). "Equations for Calculating the Rate of Leakage Migration Through
Composite Liners Due to Geomembrane Defects," Geosynthetics International, Vol. 4, Nos. 3-4,
pp. 335-348.
Giroud, J.P. and Bonaparte, R. (1989a). "Leakage Through Liners Constructed with
Geomembranes, Part 1: Geomembrane Liners," Geotextiles and Geomembranes, Vol. 8, No. 1,
pp. 27-67.
Giroud, J.P. and Bonaparte, R. (1989b). "Leakage Through Liners Constructed with
Geomembranes, Part II: Composite Liners," Geotextiles and Geomembranes, Vol. 8, No. 2, pp.
71-111.
Giroud, J.P., Khatami, A., and Badu-Tweneboah, K. (1989). "Evaluation of the Rate of Leakage
Through Composite Liners," Geotextiles and Geomembranes, Vol. 8, No. 4, pp. 337-340.
Giroud, J.P., Badu-Tweneboah, K., and Bonaparte, R. (1992). "Rate of Leakage Through a
Composite Liner Due to Geomembrane Defects," Geotextiles and Geomembranes, Vol. 11, No.
1, pp. 1-28.
Giroud, J.P., Zhao, A., and Bonaparte, R.B. (2000). "The Myth of Hydraulic Transmissivity
Equivalency Between Geosynthetic and Granular Liquid Collection Layers", Geosynthetics
International, Vol. 7, Nos. 4-6, pp. 381-401.
Gray, D.H. and Sotir, R.B. (1996). "Biotechnical and Soil Bioengineering Slope Stabilization: A
Practical Guide for Erosion Control," John Wiley & Sons, Inc., New York, NY, 400 p.
Gross, B. A., Bonaparte, R. and Giroud, J. P. (1990). "Evaluation of Flow from Landfill
Leakage Detection Layers," Proceedings of 4th International Conference on Geotextiles and
Geomembranes and Related Products, The Hague, Netherlands, pp. 481-486.
Gross, B.A., Bonaparte, R., and Othman, M.A. (1997). "Inferred Performance of Surface
Hydraulic Barriers from Landfill Operational Data," Proceedings, International Containment
Technology Conference, St. Petersburg, FL, pp. 374-380.
Gross, B.A., Bonaparte, R., and Giroud, J.P. (2002). "Waste Containment Systems: Problems
and Lessons Learned," Appendix F in Assessment and Recommendations for Optimal
DRAFT - DO NOT CITE OR QUOTE
A-7
-------
Performance of Waste Containment Systems, R. Bonaparte, D.E. Daniel, and R.M. Koerner, U.S.
Environmental Protection Agency, National Risk Management Research Laboratory, Cincinnati,
OH.
Hakonson, T.E. (1986). "Evaluation of Geologic Materials to Limit Biological Intrusion of Low-
Level Radioactive Waste Disposal Sites" LA-10286-MS, Los Alamos National Laboratory, Las
Alamos, NM.
Hansen, E.M., Schreiner, L.C., and Miller, J.F. (1982). "Application of Probable Maximum
Precipitation Estimates - United States East of the JO 5th Meridian" Hydrometerological Report
No. 52 (HMR-52), National Oceanic and Atmospheric Administration, National Weather
Service, Silver Springs, MD, 168 p.
Hanson, A.A. and Juska, F.V. (1969). "Turfgrass Science" Agronomy Monograph No. 14,
American Society of Agronomy, Madison, WI, 715 p.
Harpur, W.A., Wilson-Fahmy, R.F. and Koerner, R.M. (1994). "Evaluation of the Contact
Between GCLs and Geomembranes in Terms of Transmissivity," Proceedings, Conference on
Geosynthetic Liner Systems, R.M. Koerner and R.F. Wilson-Fahmy (eds.), Industrial Fabrics
Association International, St. Paul, MN, pp. 143-154.
Hartung, F. and Scheuerlein, H. (1970). "Design of Overflow Rockfill Dams", Proceedings of
the 10th International Conference on Large Dams, Montreal, Canada, pp. 587-598.
Hershfield, D.M. (1961). "Rainfall Intensity Atlas of the United States for Durations From 30
Minutes to 24 Hours and Return Periods from 1 to 100 Years," U.S. Department of Commerce
Weather Bureau Technical Paper No. 40, Washington D.C., 115 p.
Hewitt, R.D., and Daniel, D.E. (1997). "Hydraulic Conductivity of Geosynthetic Clay Liners
Subjected to Freeze-Thaw," Journal ofGeotechnicaland GeoenvironmentalEngineering, Vol.
123, No. 3, pp. 305-313.
Hsuan, Y.G. and Koerner, R.M. (1996). "Long-Term Durability ofHDPE Geomembranes: Part
I Depletion ofAntioxidants," GRI Report No. 16, 27 p.
Hsuan, Y.G. and Koerner, R.M. (1998). "Antioxidant Depletion Lifetime in High Density
Polyethylene Geomembranes," Journal of Geotechnical and Geoenvironmental Engineering,
ASCE, Vol. 124, No. 6, pp. 532-541.
Hsuan, Y.G. and Koerner, R.M. (2002). "Antioxidant Depletion Time in High Density
Polyethylene Geomembranes," Appendix B in "Assessment and Re commendations for Optimal
Performance of Waste Containment Systems," EPA/600/R-02/099, U.S. Environmental
Protection Agency, National Risk Management Research Laboratory, Cincinnati, OH
James, A.N., Fullerton, D., and Drake, R. (1997). "Field Performance of GCL under Ion
Exchange Conditions," Journal of Geotechnical and Geoenvironmental Engineering, Vol. 123,
No. 10, pp. 897-902.
DRAFT - DO NOT CITE OR QUOTE
A-8
-------
Johnson, D.I. and Urie, D.H. (1985). "Surface Barrier Caps: Long-Term Investments in Need of
Attention," Waste Management and Research, Vol. 3, pp. 143-148.
Kemper, W.D., Nicks, A.D., and Corey, A.T. (1994). "Accumulation of Water in Soils Under
Gravel and Sand Mulches," Soil Science Society of America Journal, Vol. 58, pp. 56-63.
Khire, M.V., Benson, C.H., and Bosscher, PJ. (1997). "Water Balance Modeling of Earthen
Final Covers," Journal of Geotechnical and GeoenvironmentalEngineering, Vol. 123, No. 8, pp.
744-754.
Khire, M.V., Benson, C.H., and Bosscher, PJ. (1999). "Field Data from a Capillary Barrier and
Model Predictions with UNSAT-H," Journal of' Geotechnical and Geoenvironmental
Engineering, Vol. 125, No. 6, pp. 518-527.
Koerner, R.M. (1998). "Designing with Geosynthetics" 4th Edition, Prentice Hall, Englewood
Cliffs, NJ, 761 p.
Koerner, R.M. and Daniel, D.E. (1997). "Final Covers for Solid Waste Landfills and Abandoned
Dumps" ASCEPress, Reston, VA, 256 p.
Koerner, G.R. and Koerner, R.M. (1995). "Temperature Behavior of Field Deployed HOPE
Geomembranes," Geosynthetics '95 Conference Proceedings, Industrial Fabrics Association
International, St. Paul, MN, Vol. 3, pp. 921-937.
Koerner, R.M. and Narejo, D. (1995). "Bearing Capacity of Hydrated Geosynthetic Clay
Liners," Journal of Geotechnical Engineering, Vol. 121, No. 1, pp. 82-85.
Koerner, R. M., Koerner, G. R. and Hwu, B.-L. (1990). "Three Dimensional, Axi-Symmetric
Geomembrane Tension Test," Geosynthetic Testing for Waste Containment Applications, ASTM
STP 1081, R.M. Koerner (ed.), American Society for Testing and Materials, Philadelphia, PA,
pp. 170-184.
Koerner, R.M., Koerner, G.R. and Eberle, M.A. (1996). "Out-of-Plane Tensile Behavior of
Geosynthetic Clay Liners," Geosynthetics International, Vol. 3, No. 2, pp. 277-296.
Kraus, J., Benson, C.H., Erickson, A.E., and Chamberlain, EJ. (1997). "Freeze-Thaw Cycling
and Hydraulic Conductivity of Bentonitic Barriers," Journal of Geotechnical and
Geoenvironmental Engineering, Vol. 123, No. 3, pp. 229-238.
LaGatta, M.D., Boardman, B.T., Cooley, B.H., and Daniel, D.E. (1997). "Geosynthetic Clay
Liners Subjected to Differential Settlement," Journal of Geotechnical and Geoenvironmental
Engineering, Vol. 123, No. 5, pp. 402-410.
Lee, C.R., Skogerboc, J.G., Eskew, K., Price, R.W., Page, N.R., Clar, M., Kort, R., and Hopkins,
H. (1984). "Restoration of Problem Soil Material at Corps of Engineers Construction Sites,"
Instruction Report EL-84-1, U.S Army Engineer Waterways Experiment Station, Vicksburg,
MS.
DRAFT - DO NOT CITE OR QUOTE
A-9
-------
Licht, L., Aitchison, E., Schnabel, W., English, M., and Kaempf, M. (2001). "Landfill Capping
with Woodland Ecosystems," Practice Journal of Hazardous, Toxic, and Radioactive Waste
Management, Vol. 5, No. 4, pp. 175-184.
Ligotke, M.W. (1994). "Control of Eolian Soil Erosion from Waste-Site Surface Barriers," In-
situ Remediation: Scientific Basis for Current and Future Technologies, G.W. Gee and N.R.
Wing (eds.), Batelle Press, Richland, WA, pp. 545-559.
Lin, L.-C. and Benson, C.H. (2000). "Effect of Wet-Dry Cycling on Swelling and Hydraulic
Conductivity of GCLs", Journal of'Geotechnical and GeoenvironmentalEngineering, Vol. 126,
No. 1, pp. 40-49.
Lyles, L. (1975). "Possible Effects of Wind Erosion on Soil Productivity," Journal of Soil and
Water Conservation, Vol. 30, pp. 279-283.
Maine Bureau of Remediation and Waste Management (1997). "An Assessment of Landfill
Cover System Barrier Layer Hydraulic Performance," Augusta, ME.
Mayne, P.W., Jones, J.S., and Dames, J.C. (1984). "Ground Response to Dynamic Compaction,"
Journal of the Geotechnical Engineering Division, Vol. 110, No. 6, pp. 757-774.
McCuen, R.H. (1998). "Hydrologic Analysis and Design" 2nd Edition, Prentice Hall Publishing
Co., NJ, 814 p.
Melchior, S. (1997a). "In-situ Studies on the Performance of Landfill Caps (Compacted Soil
Liners, Geomembranes, Geosynthetic Clay Liners, Capillary Barriers)," Proceedings,
International Containment Technology Conference, St. Petersburg, FL, pp. 365-373.
Melchior, S. (1997b). "In-situ Studies of the Performance of Landfill Caps (Compacted Soil
Liners, Geomembranes, Geosynthetic Clay Liners, Capillary Barriers)," Land Contamination
and Reclamation, Vol. 5, No. 3, pp. 209-216.
Melchior, S., Berger, K., Vielhaver, B., and Miehlich, G. (1994). "Multilayered Landfill Covers:
Field Data on the Water Balance and Liner Performance," In-Situ Remediation: Scientific Basis
for Current and Future Technologies, G.W. Gee and N.R. Wing (eds.), Battelle Press,
Columbus, OH, Part 1, pp. 411-425.
Miller, J.F., Frederick, R.H., and Tracey, RJ. (1973). "Precipitation-Frequency Atlas of the
Western United States," U.S. Department of Commerce, National Weather Service, NOAA Atlas
2, Silver Springs, MD.
Mitchell, J.K. (1993). "Fundamentals of Soil Behavior", 2nd Edition, John Wiley & Sons, Inc.,
437 p.
Mitchell, J.K. and Jaber, M. (1990). "Factors Controlling the Long-Term Properties of Clay
Liners," Waste Containment Systems: Construction, Regulation, and Performance, Geotechnical
Special Publication No. 26, R. Bonaparte (ed.), ASCE, New York, NY, pp. 84-105.
DRAFT - DO NOT CITE OR QUOTE
A-10
-------
Montgomery, RJ. and Parsons, LJ. (1989). '"Hie Omega Hills Final Cover Test Plot Study:
Three-Year Data Summary" presented at Annual Meeting of the National Solid Waste
Management Association, Washington, DC.
Montgomery, RJ. and Parsons, LJ. (1990). "The Omega Hills Final Cover Test Plot Study:
Fourth Year Data Summary" presented at 22nd Mid-Atlantic Industrial Waste Conference,
Drexel University, Philadelphia, PA.
Nelson, J.D., Abt, S.R., Volpe, R.L., VanZyl, D., Hinkle, N.E., and Staub, W.P. (1986).
"Methodologies for Evaluating Long-Term Stabilization Designs of Uranium Mill Tailings
Impoundments" NUREG/CR-4620, U.S. Nuclear Regulatory Commission, Washington, DC,
144 p.
NRC (1990). "Staff Technical Position: Design of Erosion Protection Covers for Stabilization of
Uranium Mill Tailings Sites" NRC/UMT/1003-0064, U.S. Nuclear Regulatory Commission,
Washington, DC, 23 p.
NRCS (1996). "Soil Quality Resource Concerns: Compaction" U.S. Department of Agriculture,
National Resources Conservation Service, Soil Quality Information Sheet, 2 p.
NRCS (1998a). "Soil Quality Indicators: Infiltration" U.S. Department of Agriculture, National
Resources Conservation Service, Soil Quality Information Sheet, 2 p.
NRCS (1998b). "Soil Quality Resource Concerns: Available Water Capacity" U.S. Department
of Agriculture, National Resources Conservation Service, Soil Quality Information Sheet, 2 p.
NRCS (2000). "Erosion and Sedimentation on Construction Sites" U.S. Department of
Agriculture, National Resources Conservation Service, Soil Quality Urban Technical Note No. 1,
5 p.
O'Donnell, E., Ridky, R.W., and Schultz, R.K. (1997). "Controlling Water Infiltration through
Radioactive Waste and Landfill Covers A Field Demonstration at Beltsville, Maryland,"
Proceedings, Landfill Capping in the Semi-Arid West: Problems, Perspectives, and Solutions,
T.D. Reynolds and R.C. Morris (eds.), pp. 109-128.
Othman, M.A., Benson, C.H., Chamberlain, E.J., and Zimmie, T.F. (1994). "Laboratory Testing
to Evaluate Changes in Hydraulic Conductivity of Compacted Clays Caused by Freeze-Thaw:
State-of-the-Art," Hydraulic Conductivity and Waste Containment Transport in Soils, STP 1142,
D.E. Daniel and S J. Trautwein (eds.), American Society for Testing and Materials, Philadelphia,
PA, pp. 227-254.
Othman, M.A., Bonaparte, R., Gross, B.A., and Warren, D. (2002). "Evaluation of Liquids
Management Data for Double-Lined Landfills," Appendix E in Assessment and
Recommendations for Optimal Performance of Waste Containment Systems, R. Bonaparte, D.E.
DRAFT - DO NOT CITE OR QUOTE
A-ll
-------
Daniel, and R.M. Koerner, U.S. Environmental Protection Agency, National Risk Management
Research Laboratory, Cincinnati, OH.
Rapoport, N.Ya. and Zaikov, G.E. (1986). "Kinetics and Mechanism of the Oxidation of
Stressed Polymer," in 4, Chapter 6 of Developments in Polymer Stabilization., G. Scott (ed.),
Applied Science Publishers, Ltd., London, pp. 207-258.
Renard, K.G., Foster, G.R., Weesies, G.A., McCool, O.K., and Yoder, D.C. (1997). "Predicting
Soil Loss by Water: A Guide to Conservation Planning with the Revised Universal Soil Loss
Equation (RUSLE)," Agricultural Handbook 703, U.S. Department of Agriculture, Agricultural
Research Service, 404 p.
Repa, E.W.J., Herrmann, E.F., Tokarski, E.F. and Eades, R.R. (1987). "Evaluating Asphalt Cap
Effectiveness at a Superfund Site," Journal of Environmental Engineer ing, Vol. 113, No. 3, pp.
649-653.
Richards, L.A. (1931). "Capillary Conduction of Liquids in Porous Mediums," Physics, Vol. 1,
pp. 318-333.
Rogers and Associates Engineering (1984a). "Radon Attenuation Effectiveness and Cover
Optimization with Moisture Effects (RAECOM)," computer program, U.S. Nuclear Regulatory
Commission, Washington, DC.
Rogers and Associates Engineering (1984b). "Radon Attenuation Handbook for Uranium Mill
Tailings Cover System" NUREG/CR-3533, U.S. Nuclear Regulatory Commission, Washington,
DC.
Rowe, R.K. (1998). "Geosynthetics and the Minimization of Contaminant Migration through
Barrier Systems beneath Solid Waste," Proceedings, Sixth International Conference on
Geosynthetics, Industrial Fabrics Association International, St. Paul, MN, Vol. 1, pp. 27-102.
Ruhl, J.L. and Daniel, D.E. (1997). "Geosynthetic Clay Liners Permeated with Chemical
Solutions and Leachates," Journal of'Geotechnical and GeoenvironmentalEngineering., Vol.
123, No. 4, pp. 369-381.
Scanlon, B.R., Nicot, J.P., and Massmann, J.M. (1999). "Soil Gas Movement in Unsaturated
Systems," Handbook of Soil Science, M.E. Sumner (ed.), CRC Press, Boca Raton, FL, pp. A277-
A319.
Schroeder, P.R., Dozier, T.S., Zappi, P.A., McEnroe, B.M., Sjostrom, J.W., and Peyton, R. L.
(1994). The Hydrologic Evaluation of Landfill Performance (HELP) Model: Engineering
Documentation for Version 3, EPA/600/R-94/168b, U.S. Environmental Protection Agency Risk
Reduction Engineering Laboratory, Cincinnati, OH, 116p.
SCS (1986a). "Urban Hydrology for Small Watersheds: Technical Release 55," U.S.
Department of Agriculture, Soil Conservation Service.
DRAFT - DO NOT CITE OR QUOTE
A-12
-------
SCS (1986b). "Engineering Manual for Field Conservation Practices" PB85-175164, U. S.
Department of Agriculture, Soil Conservation Service.
Shan, H.-Y. and Daniel, D.E. (1991). "Results of Laboratory Tests on a Geotextile/Bentonite
Liner Material", Geosynthetics '91 Conference Proceedings., Industrial Fabrics Association
International, St. Paul, MN, pp. 517-535.
Sherard, J.L., Dunnigan, L.P., and Decker, R.S. (1976). "Identification and Nature of Dispersive
Soils," Journal of the Geotechnical Engineering Division, Vol. 102, No. GT4, pp. 287-301.
Steiniger, F. (1968). "The Effect of Burrower Attack on Dike Liners," Wasser undBoden,
Berlin.
Stephenson, D. (1979). "Rockftll in Hydraulic Engineering" Elsevier Scientific Publishing
Company, New York, NY.
Steurer, P. (1998). "Data Documentation for Air-Freezing Index Statistics for the United
States" TD-9712d, National Climatic Data Center, Asheville, NC.
Suter, II, G.W., Luxmoore, R.J., and Smith, E.D. (1993). "Compacted Soil Barriers at
Abandoned Landfill Sites Are Likely to Fail in the Long Term," Journal of Environmental
Quality, Vol. 22, pp. 217-226.
Swope, G. L. (1975). "Revegetation of Landfill Cover Sites," M.S. Thesis, Pennsylvania State
University, State Park, PA.
Temple, D.M., Robinson, K.M., Ahring, R.M., and Davis, A.G. (1987). "Stability Design of
Grass-Lined Open Channels" Agriculture Handbook No. 667, U.S. Department of Agriculture,
Agriculture Research Service, Beltsville, MD, 167 p.
Theisen, M.S. (1992). "The Role of Geosynthetics in Erosion and Sediment Control: An
Overview," Journal of Geotextiles and Geomembranes, Vol. 11, Nos. 4-6, pp. 199-214.
Thornburg, A.A. (1979). "Plant Materials for Use on Surface Mined Lands" TP-157, EPA-
600/7-79-134, U. S. Department of Agriculture, Soil Conservation Service, Washington, DC.
Vanoni, V.A. (ed.) (1975). "Sedimentation Engineering" ASCE Task Committee for the
Preparation of the Manual on Sedimentation, ASCE, New York, NY, 745 p.
Viebke, I, Elble, E., and Gedde, U.W. (1994). "Degradation of Unstabilized Medium Density
Polyethylene Pipes in Hot Water Applications," Polymer Engineering and Science, Vol. 34, No.
17, pp. 1354-1361.
Wagner, L.E. (1996). "Wind Erosion Prediction System (WEPS): Overview," USDA-ARS
Contribution No. 96-205-A.
DRAFT - DO NOT CITE OR QUOTE
A-13
-------
Waugh, WJ. (1994). "Paleoclimatic Data Application: Long-Term Performance of Uranium
Mill Tailings Repositories," Workshop Proceedings: Climate Change in the Four Corners and
Adjacent Regions, Grand Junction, CO, Sept 12-14.
Wong, L.C., and Haug, M.D. (1991). "Cyclical Closed-System Freeze-Thaw Permeability
Testing of Soil Liner and Cover Materials," Canadian GeotechnicalJournal, Vol. 28, pp. 784-
793.
Woodruff, N.P., and Siddoway, F.H. (1965). "A Wind Erosion Equation," Soil Sci. Soc. Am.
Proc. 29(5), pp. 602-608.
Wright, MJ. (ed.) (1976). "Plant Adaptation to Mineral Stress in Problem Soils," Cornell
University Agricultural Experiment Station, Ithaca, NY.
Chapter 3
Anderson, I.E., Nowak, R.S., Ratzlaff, T.D., and Markham, O.D. (1993). "Managing Soil
Moisture on Waste Burial Sites in Arid Regions" Journal of Environmental Quality, Vol. 22, pp.
62-69.
Ahmed, I. (1993). "Laboratory Study on Properties of Rubber Soils" FHWA/IN/JHRP-93/4,
School of Civil Engineering, Purdue University, West Layfayette, IN, 348 p.
Benson, C.H. (1997). "A Review of Alternative Landfill Cover Demonstrations," Environmental
Geotechnics Report No. 97-1, Geotechnical Engineering Program, Department of Civil &
Environmental Engineering, University of Wisconsin, Madison, WI.
Birkeland, P.W. (1984). "Soils and Geomorphology", Oxford University Press, New York, NY,
pp. 138-140.
Caldwell, J.A. and Reith, C.C. (1993). "Principles and Practice of Waste Encapsulation" Lewis
Publishers, Chelsea, MI, 414 p.
Cecich, V., Gonzales, L., Hoisaeter, A., Williams, J., and Reddy, K. (1996). "Use of Shredded
Tires as Lightweight Backfill Material for Retaining Structures", Waste Management &
Research, No. 14, pp. 433-451.
Cheng, S.C.J., Corcoran, G.T., Miller, C.J., and Lee, J.Y. (1994). "The Use of a Spray
Elastomer for Landfill Cover Applications," Proceedings of the Fifth International Conference
on Geotextiles, Geomembranes, and Related Products, Singapore, pp. 1037-1040.
Duffy, D.P. (1996). "The Potential Use of Shredded Tire Chips as a Leachate Drainage and
Collection Medium: Design, Construction, and Performance Considerations," Landfill Design
Using Tire Shreds, Texas Natural Resource Conservation Commission Seminar, Arlington, TX,
pp.113-132.
Dwyer, S.F. (1997). "Large-Scale Field Study of Landfill Covers at Sandia National
Laboratories," Conference Proceedings: Landfill Capping in the Semi-Arid West: Problems,
DRAFT - DO NOT CITE OR QUOTE
A-14
-------
Perspectives, and Solutions., T.D. Reynolds and R.C. Morris (eds.), Grand Teton National Park,
Wyoming, Sandia National Laboratories, Albuquerque, NM, pp. 87-107.
Dwyer, S.F. (1998). "Alternative Landfill Covers Pass the Test", Civil Engineering, ASCE, Vol.
68, No. 9, pp. 50-52.
Dwyer, Stephen F., J. Stormont,, and C. Anderson. (1999). "Mixed Waste Landfill Design
Report," SandiaNational Laboratory Report, SAND99-2514.
Dwyer, S.F. (2001). "Finding a Better Cover," Civil Engineering, ASCE, Vol. 71, No. 1, pp 58-
63.
Edil, T.B., Fox, P.J., and AM, S.W. (1992). "Hydraulic Conductivity and Compressibility of
Waste Tire Chips," Proceedings, Fifth Annual Madison Waste Conference, University of
Wisconsin, Madison, WI, pp. 49-61.
EPA (1989). "Final Covers on Hazardous Waste Landfills and Surface Impoundments"
Technical Guidance Document, EPA/530/SW-89/047, U.S. Environmental Protection Agency,
Office of Solid Waste and Emergency Response, Washington, D.C., 39 p.
EPA (1991). "Design and Construction of RCRA/CERCLA Final Covers" Seminar Publication,
EPA/625/4-91/025, U.S. Environmental Protection Agency, Office of Research and
Development, Washington, D.C., 145 p.
EPA (2003), "Evapotranspiration Landfill Cover Systems Fact Sheet, " EPA 542-F-03-015,
U.S. Environmental Protection Agency.
Payer, MJ. and Jones, T.L. (1990). "UNSAT-H, Version 2.0: Unsaturated Soil Water and Heat
Flow Model" PNL-6779, Pacific Northwest Laboratory, Richland, WA.
Gasper, AJ. (1990). "Stabilized Foam as Landfill Daily Cover," Proceedings, Municipal Solid
Waste Management: Solutions for the 90's, U.S. Environmental Protection Agency, Washington,
D.C., pp. 1113-1121.
GeoSyntec Consultants (1998a). "Test Pad Demonstration Program - Tire Shreds as Cover
Foundation, Leachate Drainage, and Operations Layer Material at Municipal Solid Waste
Landfills" report for California Integrated Waste Management Board.
GeoSyntec Consultants (1998b). "Guidance Manual - Tire Shreds as Gas Collection Material
at Municipal Solid Waste Landfills" report for California Integrated Waste Management Board.
GeoSyntec Consultants (1998c). "Guidance Manual - Tire Shreds as Final Cover Foundation
Layer Material at Municipal Solid Waste Landfills" report for California Integrated Waste
Management Board.
Glade, M.J., Bollman, D.D., and Tagawa, L.T. (1993). "Beneficial Use of Shredded Tires at the
Lowery Landfill Superfund Site: A Case
Conference, Madison, WI, pp. 442-459.
Lowery Landfill Superfund Site: A Case Study," Proceedings, 16th International Madison Waste
DRAFT - DO NOT CITE OR QUOTE
A-15
-------
Hakonson, I.E., Lane, L.G., and Springer, E.P. (1992). "Biotic and Abiotic Processes," Chapter
4 in Deserts as Dumps?, C.C. Reith and B.M. Thomson (eds.), University of New Mexico Press.
Hakonson, T.E., Maines, K.L., Warren, R.W., Bostick, K.V., Trujillo, G., Kent, J.S., and Lane,
LJ. (1994). "Hydrologic Evaluation of Four Landfill Cover Designs at Hill Air Force Base,
Utah" Report LAUR-93-4469, Los Alamos National Laboratory, Los Alamos, NM.
Hauser, V.L., Shaw, M.A., and Weand, B.L. (1994). "Effectiveness of Soil-Vegetative Covers
for Waste Sites," SUPERFUND XV: 15th Environmental Conference and Exhibition for the
Hazardous Materials/Hazardous Waste Management Industry, Washington DC, pp. 368-374.
Hillel, D. and Baker, R.S. (1988). "A Descriptive Theory of Fingering During Infiltration into
Layered Soils," Soil Science, Vol. 146, pp. 51-56.
Hillel, D. (1998). "Environmental Soil Physics" Academic Press, San Diego, CA, 771 pp.
Horvath, J.S. (1995a). "Geofoam Geosynthetic," HorvathEngineering, P.C., Scarsdale, N.Y.,
229 p.
Horvath, J.S. (1995b). "EPS Geofoam: New Products and Marketing Trends" Geotechnical
Fabrics Report, Vol. 13, No. 6, Industrial Fabrics Association International, St. Paul, MN, pp.
22-26.
Humphrey, D.N. (1996). "Investigation of Exothermic Reaction in Tire Shred Fill Located on
SR 100 in Ilwaco, Washington", Federal Highway Administration, Washington, D.C., 44 p.
Humphrey, D.N., Sanford, T.C, Cribbs, M.M., and Manion, W.P. (1993). "Shear Strength and
Compressibility of Tire Chips for Use as Retaining Wall Backfill", Transportation Research
Record 1422, Transportation Research Board, Washington, D.C., pp. 29-35.
Humphrey, D.N., Katz, L.E., and Blumenthal, M. (1997). "Water Quality Effects of Tire Chip
Fills Placed Above the Groundwater Table," Testing Soil Mixed with Waste or Recycled
Materials, ASTM STP 1275, M.A. Wasemiller and K.B. Hoddinott (eds.), American Society for
Testing and Materials, Philadelphia, PA, pp. 299-313.
ITRC (2003). "Technical and Regulatory Guidance for Design, Installation, and Monitoring of
Alternative Final Covers", Interstate Technology and Regulatory Council, Washington, D.C.,
125 p.
Khire, M. V. (1995). "Field Hydrology and Water Balance Modeling of Earthen Final Covers
for Waste Containment," Environmental Geotechnics Report No. 95-5, Geotechnical
Engineering Program, Department of Civil & Environmental Engineering, University of
Wisconsin, Madison, WI, 166 p.
Khire, M.V., Benson, C.H., and Bosscher, PJ. (2000). "Capillary Barriers: Design Variables and
the Water Balance," Journal of Geotechnical and Geoenvironmental Engineering, ASCE, Vol.
126,No.8, pp. 695-708.
DRAFT - DO NOT CITE OR QUOTE
A-16
-------
Koerner, R.M. and Daniel, D.E. (1997). "Final Covers for Solid Waste Landfills and Abandoned
Dumps" ASCEPress, Reston, VA, 256 p.
Licht, L., Aitchison, E., Schnabel, W., English, M., Kaempf, M. (2001). "Landfill Capping with
Woodland Ecosystems," Practice Journal of Hazardous, Toxic, and Radioactive Waste
Management, Vol. 5, No. 4, pp. 175-184.
Lincoln, R.J., Boxshall, G.A., and Clark, P.P. (1982). "A Dictionary of Ecology, Evolution, and
Systematics," Cambridge University Press, Cambridge, 298 p.
Meigs, P. (1953). "World Distribution of Arid and Semi-arid Homoclimates," Reviews of
Research on Arid Zone Hydrology, United Nations Educational, Scientific, and Cultural
Organization, Arid Zone Programme-1, Paris, pp. 203-209.
Miller, C.J., Cheng, S.C.J., and Tanis, R. (1997). "Poly-Urea Spray Elastomer for Waste
Containment Applications," Proceedings, International Containment Technology Conference,
St. Petersburg, FL, pp. 334-340.
Moo-Young H.K. and Zimmie T.F. (1996). "Geotechnical Properties of Paper Mill Sludges for
Use in Landfill Covers," Journal of Geotechnical Engineering, Vol. 122, No. 9, pp. 768-776.
Morris, C.E. and Stormont, J.C. (1997a). "Numerical Simulations of Capillary Barrier Field
Tests," Proceedings, International Containment Technology Conference, St. Petersburg, FL, pp.
275-281.
Morris, C.E. and Stormont, J.C. (1997b). "Capillary Barriers and Subtitle D Covers: Estimating
Equivalency," Journal of Environmental Engineering, Vol. 123, No. 1, pp. 3-10.
Nickels, W.L., Jr. (1995). "The Effect of Tire Chips as Subgrade Fill on Paved Roads," M.S.
Thesis, Department of Civil Engineering, University of Maine, Orono, ME, 215 p.
Nyhan, J.W., Hakonson, T.E., and Drennon, B.J. (1990). "A Water Balance Study of Two
Landfill Cover Designs for Semi-Arid Regions," Journal of Environmental Quality, Vol. 19, No.
2, pp. 281-288.
Nyhan, J.W., Schofield, T.G., and Starmer, R.H. (1997). "Water Balance Study of Four Landfill
Cover Designs Varying in Slope for Semi-Arid Regions," Journal of Environmental Quality,
Vol. 26, No. 5, pp. 1385-1392.
Penman, H.L. (1948). "Natural Evaporation From Open, Bare Soil and Grass." Proceedings
Royal Society of London, A193, pp. 120-145
Richards, L.A. (1931). "Capillary Conduction of Liquids Through Porous Media", Physics, Vol.
1, pp. 318-333.
Schroeder, P.R., Lloyd, C.M., and Zappi, P.A. (1994a). "The Hydrologic Evaluation of Landfill
Performance (HELP) Model: User's Guide for Version 3," EPA/600/R-94/168a, U.S.
DRAFT - DO NOT CITE OR QUOTE
A-17
-------
Environmental Protection Agency, Risk Reduction Engineering Laboratory, Cincinnati, OH, 84
P-
Schroeder, P.R., Dozier, T.S., Zappi, P.A., McEnroe, B.M., Sjostrom, J.W., and Peyton, R. L.
(1994b). " The Hydrologic Evaluation of Landfill Performance (HELP) Model: Engineering
Documentation for Version 3," EPA/600/R-94/168b, U.S. Environmental Protection Agency
Risk Reduction Engineering Laboratory, Cincinnati, OH, 116 p.
Stormont, J.C. (1995). "The Performance of Two Capillary Barriers During Constant
Infiltration," Landfill Closures... Environmental Protection and Land Recovery, Geotechnical
Special Publication No. 53, R. J. Dunn and U.P. Singh (eds.), ASCE, New York, NY, pp. 77-92.
Stormont, J.C. (1996). "The Effectiveness of Two Capillary Barriers on a 10% Grade,"
Geotechnical and Geological Engineering, Vol. 14, pp. 243-267.
Stormont, J.C. (1997). "Incorporating Capillary Barriers in Surface Cover Systems," Conference
Proceedings - Landfill Capping in the Semi-Arid West: Problems, Perspectives, and Solutions,
T.D. Reynolds and R.C. Morris (eds.), Grand Teton National Park, Wyoming, pp. 39-51.
Stormont, J.C. and Morris, C.E. (1998). "Method to Estimate Water Storage Capacity of
Capillary Barriers," Journal of Geotechnical and Geoenvironmental Engineering, Vol. 124, No.
4, pp. 297-302.
Stormont, J.C., Henry, K.S., and Evans, T.M. (1997). "Water Retention Functions of Four
Nonwoven Polypropylene Geotextiles," Geosynthetics International, Vol. 4, No. 6, pp. 661-672.
Suter, II, G.W., Luxmoore, R.J., and Smith, E.D. (1993). "Compacted Soil Barriers at
Abandoned Landfill Sites Are Likely to Fail in the Long Term," Journal of Environmental
Quality, Vol. 22, pp. 217-226.
Waugh, W.J. and Petersen, K.L. (1995). "Paleoclimatic Data Application: Long-Term
Performance of Uranium Mill Tailings Repositories," Workshop Proceedings: Climate Change
in the Four Corners and Adjacent Regions: Implications for Environmental Restoration and
Land-Use Planning, W.J. Waugh, K.L. Petersen, P.E. Wigand, and B. Louthan (eds.), Grand
Junction, CO.
White, R. (1995a). "EPS Geofoam: Unique Solutions to Forming Steep Landfill Embankments,"
Geosynthetics World, Geosynthetics Publications, Ltd., U.K., Vol. 5, No. 2, p. 10.
White, R. (1995b). "EPS Used to Assist in Methane and Radon Gas Venting," Geosynthetics
World, Geosynthetics Publications, Ltd., U.K., Vol. 5, No. 2, p. 12.
Zimmie T.F. and Moo-Young H.K. (1995). "Hydraulic Conductivity of Paper Sludges Used for
Landfill Covers," Geoenvironment 2000, Geotechnical Special Publication No. 46, D.E. Daniel
and Y.B. Acar (eds.), ASCE, New York, NY, Vol. 2, pp. 932-946.
Chapter 4
DRAFT - DO NOT CITE OR QUOTE
A-18
-------
Anderson, I.E. (1997). "An Ecological Engineering Approach for Keeping Water from
Reaching Interred Wastes in Arid or Semi-Arid Regions," Proceedings, International
Containment Technology Conference, St. Petersburg, FL, pp. 243-248.
Anderson, I.E., Nowak, R.S., Ratzlaff, T.D., and Markham, O.D. (1993). "Managing Soil
Moisture on Waste Burial Sites in Arid Regions" Journal of Environmental Quality, Vol. 22, pp.
62-69.
Arnold, J.G., Williams, J.R., Nicks, A.D., and Sammons, N.B. (1989). "SWRRB, a Simulator for
Water Resources in Rural Basins" U.S. Department of Agriculture, Agricultural Research
Service, Texas A&M University Press, College Station, TX.
Barnes, FJ. and Rodgers, I.E. (1988). "Evaluation of Hydrologic Models in the Design of Stable
Landfill Covers" EPA Project Summary, EPA/600/S2-88/048, U.S. Environmental Protection
Agency, Office of Research and Development, Washington, D.C.
Benson, C.H., Khire, M.V., and Bosscher, PJ. (1993). "Final Cover Hydrologic Evaluation,
Final Cover - Phase II" Environmental Geotechnics Report No. 93-4, Geotechnical Engineering
Program, Department of Civil & Environmental Engineering, University of Wisconsin, Madison,
WI, 151 p.
Campbell, G.S. (1974). "A Simple Method for Determining Unsaturated Hydraulic Conductivity
from Moisture Retention Data," Soil Science, Vol. 117, No. 6, pp. 311-314.
Carroll, R.G. (1983). "Geotextile Filter Criteria," Engineering Fabrics in Transportation
Construction, Transportation Research Record 96, Transportation Research Board, Washington,
D.C., pp. 46-53.
Cartwright, K., Larson, T.H., Herzog, B.L., Johnson, T.M., Albrecht, K.A., Moffett, D.L.,
Keffer, D.A., and Stohr, CJ. (1988). "Trench Covers to Minimize Infiltration at Waste Disposal
Sites" Circular 541, Illinois State Geological Survey, Champaign, 111., Circular 541, 88 p.
Cedergren, H.R. (1989). Seepage, Drainage, and Flow Nets, 3rd Edition, John Wiley & Sons,
Inc., New York, NY, 464 p.
Childs, S.W. and Hanks, R.J. (1975). "Model of Soil Salinity Effects on Crop Growth," Soil
Science Society of American Journal, Vol. 39, pp. 617-622.
Dwyer, S.F. (1995). "Alternative Landfill Cover Demonstration," Landfill Closures...
Environmental Protection and Land Recovery, Geotechnical Special Publication No. 53, R.J.
Dunn and U.P. Singh (eds.), ASCE, New York, NY, pp. 19-34.
Dwyer, S.F. (1998). "Alternative Landfill Covers Pass the Test," Civil Engineering, ASCE, Vol.
68, No. 9, pp. 50-52.
Dwyer, S.F. (2001). "Finding a Better Cover," Civil Engineering, ASCE, Vol. 71, No. 1, pp 58-
63.
DRAFT - DO NOT CITE OR QUOTE
A-19
-------
EPA (1991). "Design and Construction of RCRA/CERCLA Final Covers" Seminar Publication,
EPA/625/4-91/025, U.S. Environmental Protection Agency, Office of Research and
Development, Washington, D.C., 145 p.
Estornell, P. and Daniel, D.E. (1992). "Hydraulic Conductivity of Three Geosynthetic Clay
Liners," Journal of Geotechnical Engineering, Vol. 118, No. 10, pp. 1592-1606.
Payer, M.J., and Jones, T.L. (1990). "UNSAT-H, Version 2.0: Unsaturated Soil Water and Heat
Flow Model" PNL-6779, Pacific Northwest Laboratory, Richland, WA.
Payer, M.J., Rockhold, M.L., and Campbell, M.D. (1992). "Hydrologic Modeling of Protective
Barriers: Comparison of Field Data and Simulation Results," Soil Science Society of America
Journal, Vol. 56, No. 3, pp. 690-700.
Fenn, D.G., Hanley, K.J., and DeGeare, T.V. (1975). "Use of the Water Balance Method for
Predicting Leachate Generation from Solid Waste Disposal Sites" EPA/530/SW-168, U.S.
Environmental Protection Agency, Washington, D.C., 40 p.
Fleenor, W.E. and King, IP. (1995). "Identifying Limitations on Use of the HELP Model,"
Landfill Closures...Environmental Protection and Land Recovery, Geotechnical Special
Publication No. 53, RJ. Dunn and U.P. Singh (eds.), ASCE, New York, NY, pp. 121-138.
Fredlund, D.G., and Xing, A. (1994). "Equations for the Soil-Water Characteristic Curve,"
Canadian Geotechnical Journal, Vol. 31, pp. 521-532.
Fredlund, D.G., Xing, A., and Huang, S. (1994). "Predicting the Permeability Function for
Unsaturated Soils Using the Soil-Water Characteristic Curve" Canadian Geotechnical Journal,
Vol. 31, pp. 533-546.
Gee, G.W., Wierenga, P.J., Andraski, B.J., Young, M.H., Payer, M.J., and Rockhold, M.L.
(1994). "Variations in Water Balance and Recharge Potential at Three Western Desert Sites,"
Soil Science Society of America Journal, Vol. 58, pp. 63-72.
Gee, G.W., Anderson, L.W., and Payer, MJ. (1997). "Surface Barrier Research at the Hanford
Site," Proceedings, International Containment Technology Conference, St. Petersburg, PL, pp.
305-311.
Giroud, J.P. (1982). "Filter Criteria for Geotextiles," Proceedings, Second International
Conference on Geotextiles, Vol. 1, Las Vegas, NV, pp. 103-108.
Giroud, J.P. (1996). "Granular Filters and Geotextile Filters," Proceedings, GeoFilters '96,
Montreal, Canada, pp. 565-680.
Giroud, J.P. and Bonaparte, R. (1989a). "Leakage Through Liners Constructed with
Geomembranes, Part 1: Geomembrane Liners," Geotextiles andGeomembranes, Vol. 8, No. 1,
pp. 27-67.
DRAFT - DO NOT CITE OR QUOTE
A-20
-------
Giroud, J.P. and Bonaparte, R. (1989b). "Leakage Through Liners Constructed with
Geomembranes, Part II: Composite Liners," Geotextiles and Geomembranes, Vol. 8, No. 2, pp.
71-111.
Giroud, J.P., Badu-Tweneboah, K., and Bonaparte, R. (1992a). "Rate of Leakage Through a
Composite Liner Due to Geomembrane Defects," Geotextiles and Geomembranes., Vol. 11, No.
1, pp. 1-28.
Giroud, J.P., Gross, B.A., and Darasse, J. (1992b). "Flow in Leachate Collection Layers -
Steady-State Conditions," GeoSyntec Consultants Internal Report, Boca Raton, FL, 62 p.
Giroud, J.P. and Houlihan, M.F. (1995). "Design of Leachate Collection Layers," Proceedings,
Fifth International Landfill Symposium, Sardinia, Italy, pp. 613-640.
Gross, B.A., Bonaparte, R., and Othman, M.A. (1997). "Inferred Performance of Surface
Hydraulic Barriers from Landfill Operational Data," Proceedings, International Containment
Technology Conference, St. Petersburg, FL, pp. 374-380.
Gupta S.K., Tanji, K.K., Nielsen, D.R., Biggar, J.W., Simmons, C.S., and Maclntyre, J.L.
(1978). "Field Simulation of Soil-Water Movement with Crop Water Extraction," Water Science
and Engineering Paper No. 4013, Department of Land, Air, and Water Resources, University of
California, Davis, CA.
Hutson, J.L. and Wagenet, R.J. (1992). "LEACHMLeaching Estimation and Chemistry Model,"
Version 3, Department of Soil Crop and Atmospheric Sciences Research Series No. 92-3, New
York State College of Agricultural and Life Sciences, Cornell University, Ithaca, NY.
Karr, L.A., Harre, B., and Hakonson, T.E. (1997). "Alternative Landfill Cover Technology
Demonstration at Kaneohe Marine Corps Base Hawaii," Proceedings, International Containment
Technology Conference, St. Petersburg, FL, pp. 407-413.
Khire, M. V. (1995). "Field Hydrology and Water Balance Modeling of Earthen Final Covers
for Waste Containment," Environmental Geotechnics Report No. 95-5, Geotechnical
Engineering Program, Department of Civil & Environmental Engineering, University of
Wisconsin, Madison, WI, 166 p.
Khire, M.V., Benson, C.H., and Bosscher, P.J. (1997). "Water Balance Modeling of Earthen
Final Covers," Journal of Geotechnical and Geoenvironmental Engineering, Vol. 123, No. 8, pp.
744-754.
Khire, M.V., Benson, C.H., and Bosscher, P.J. (1999). "Field Data from a Capillary Barrier and
Model Predictions with UNSAT-H," Journal of Geotechnical and Geoenvironmental
Engineering, Vol. 125, No. 6, pp. 518-527.
Kmet, P. (1982). "EPA Water Balance Method - Its Use and Limitations," Wisconsin
Department of Natural Resources, Madison, WI.
DRAFT - DO NOT CITE OR QUOTE
A-21
-------
Koerner, R.M. (1998). "Designing with Geosynthetics," 4th Edition, Prentice Hall, Englewood
Cliffs, NJ, 761 p.
Koerner, R.M. and Daniel, D.E. (1997). "Final Covers for Solid Waste Landfills and Abandoned
Dumps" ASCEPress, Reston, VA, 256 p.
Lafleur, J., Mlynarek, J., and Rollin, A.L. (1989). "Filtration of Broadly Graded Cohesionless
Soils," Journal of GeotechnicalEngineering, Vol. 115, No. 12, pp. 1747-1768.
Lane, D.T., Benson, C.H., and Bosscher, PJ. (1992). "Hydrologic Observations and Modeling
Assessments of Landfill Covers," Environmental Geotechnics Report No. 92-10, Geotechnical
Engineering Program, Department of Civil & Environmental Engineering, University of
Wisconsin, Madison, WI, 406 p.
Laundre, J.W. (1997). "The Impact of a Shallow Biobarrier on Water Recharge Patterns in a
Semi-Arid Environment," Proceedings, International Containment Technology Conference., St.
Petersburg, FL, pp. 243-248.
Luettich, S.M., Giroud, J.P., and Bachus, R.C. (1992). "Geotextile Filter Design Guide,"
Geotextiles and Geomembranes, Vol. 11, No. 4-6, pp. 355-370.
Limbach, W.E., Ratzlaff, T.D., Anderson, I.E., Reynolds, T.D., and Laundre, J.W. (1994).
"Design and Implementation of the Protective Cap/Biobairier Experiment at the Idaho National
Engineering Laboratory," In-Situ Remediation: Scientific Basis for Current and Future
Technologies, G.W. Gee andN.R. Wing, (eds.). Batelle Press, Richland, Wash., pp. 359-377.
Martian, P. (1994). Calibration of HELP Version 2.0 and Performance Assessment of Three
Infiltration Barrier Designs for HanfordSite Remediation, Draft A, DOE/RL-93-33, Idaho
National Engineering Laboratory, Idaho Falls, Idaho, 69 p.
McBean, E.A., Rovers, F.A., and Farquhar, GJ. (1995). "Solid Waste Landfill Engineering and
Design," Prentice Hall PTR, Englewood Cliffs, N.J., 521 p.
McEnroe, B.M. (1993). "Maximum Saturated Depth Over Landfill Liner," Journal of
Environmental Engineering, Vol. 119, No. 2, pp. 262-270.
Melchior, S. (1997a). "In-situ Studies on the Performance of Landfill Caps (Compacted Soil
Liners, Geomembranes, Geosynthetic Clay Liners, Capillary Barriers)," Proceedings,
International Containment Technology Conference, St. Petersburg, FL, pp. 365-373.
Melchior, S. (1997b). "In-situ Studies of the Performance of Landfill Caps (Compacted Soil
Liners, Geomembranes, Geosynthetic Clay Liners, Capillary Barriers)," Land Contamination
and Reclamation, Vol. 5, No. 3, pp. 209-216.
Melchior, S., Berger, K., Vielhaver, B., and Miehlich, G. (1994). "Multilayered Landfill Covers:
Field Data on the Water Balance and Liner Performance," In-Situ Remediation: Scientific Basis
for Current and Future Technologies, G.W. Gee andN.R. Wing (eds.), Battelle Press,
Columbus, OH, Part 1, pp. 411-425.
DRAFT - DO NOT CITE OR QUOTE
A-22
-------
Morris, C.E. and Stormont, J.C. (1997). "Numerical Simulations of Capillary Barrier Field
Tests," Proceedings, International Containment Technology Conference, St. Petersburg, FL, pp.
275-281.
Nichols, W.E. (1991). "Comparative Simulations of a Two-Layer Landfill Barrier Using the
HELP Version 2.0 and UNSAT-H Version 2.0 Computer Codes" PNL-7583, Pacific Northwest
Laboratory, Richland, WA.
Nyhan, J.W. (1989). Hydrologic Modeling to Predict Performance of Shallow LandBurial
Cover Designs at the Los Alamos National Laboratory, LA-11533-MS, Los Alamos National
Laboratory, Los Alamos, NM, 34 p.
Nyhan, J.W., Hakonson, I.E., and Drennon, BJ. (1990). "A Water Balance Study of Two
Landfill Cover Designs for Semi-Arid Regions," Journal of Environmental Quality, Vol. 19, No.
2, pp. 281-288.
Nyhan, J.W., Schofield, T.G., and Starmer, R.H. (1997). "Water Balance Study of Four Landfill
Cover Designs Varying in Slope for Semi-Arid Regions," Journal of Environmental Quality,
Vol. 26, No. 5, pp. 1385-1392.
Paige, G.B., Stone, J.J., Lane, L.J., and Hakonson, I.E. (1996). "Calibration and Testing of
Simulation Models for Evaluation of Trench Cap Designs," Journal of Environmental Quality,
Vol. 25, pp. 136-144.
Peters, N., Warner, R.S., Coates, A.L., Logsdon, D.S., and Grube, W.E. (1986). "Applicability
of the HELP Model in Multilayer Cover Design: A Field Verification and Modeling
Assessment," Land Disposal of Hazardous Waste - Proceedings, 1986 Research Symposium,
U.S. Environmental Protection Agency, Cincinnati, OH.
Peyton, R.L. and Schroeder, P.R. (1988). "Field Verification of HELP Model for Landfills,"
Journal of Environmental Engineering, ASCE, Vol. 114, No. 2, pp. 247-269.
Peyton, R.L. and Schroeder, P.R. (1993). "Water Balance for Landfills," GeotechnicalPractice
for Waste Disposal, D.E. Daniel (ed.). Chapman & Hall, London, England, pp. 214-243.
Richards, L.A. (1931). "Capillary Conduction of Liquids in Porous Mediums," Physics, Vol. 1,
pp. 318-333.
Richardson, C.W. and Wright, D.A. (1984). "WGEN: A Model for Generating Daily Weather
Variables," ARS-8, U.S. Department of Agriculture, Agricultural Research Service, 83 p.
Ritchie, J.T. (1972). "Model for Predicting Evaporation From a Row Crop with Incomplete
Cover," Water Resources Research, Vol. 8, No. 5, pp. 1204-1213.
Sackschewsky, M.R., Kemp, C.J., Link, S.O., and Waugh, W.J. (1995). "Soil Water Balance
Changes in Engineered Soil Surfaces," Journal of Environmental Quality, Vol. 24, pp. 352-359.
DRAFT - DO NOT CITE OR QUOTE
A-23
-------
Scanlon, B.R., Christman, M.C., Reedy, R.C., Porro, I, Simunek, I, Flerchinger, G.N. (2002)
"Intercede Comparisons for Simulating Water Balance of Surficial Sediments in Semiarid
Regions," Water Resources Research (in review).
Scharch, P.E. (1985). "Water Balance Analysis Program for the IBM-PC Microcomputer'"
Wisconsin Department of Natural Resources, Madison, WI, 1985.
Schroeder, P.R., Lloyd, C.M., and Zappi, P.A. (1994a). The Hydrologic Evaluation of Landfill
Performance (HELP) Model: User's Guide for Version 3, EPA/600/R-94/168a, U.S.
Environmental Protection Agency, Risk Reduction Engineering Laboratory, Cincinnati, OH.
Schroeder, P.R., Dozier, T.S., Zappi, P.A., McEnroe, B.M., Sjostrom, J.W., and Peyton, R. L.
(1994b). The Hydrologic Evaluation of Landfill Performance (HELP) Model: Engineering
Documentation for Version 3, EPA/600/R-94/168b, U.S. Environmental Protection Agency Risk
Reduction Engineering Laboratory, Cincinnati, OH, 116p.
Schultz, R.K., Ridky, R.W., and O'Donnell, E. (1995). "Control of Water Infiltration into Near
Surface LLW Disposal Units, Progress Report of Field Experiments at a Humid Region Site,
Beltsville, Maryland" NUREG/CR-4918, U.S. Nuclear Regulatory Commission, Washington,
DC, Vol. 8, 20 p.
SCS (1985). User's Guide for the CREAMS Computer Model, Washington Computer Center
Version, Technical Release 72.
Simunek, J., Sejna, M., and van Genuchten, M.Th. (1999). "The HYDRUS-2D Software
Package for Simulating the Two-Dimensional Movement of Water, Heat and Multiple Solutes in
Variably-Saturated Media, Version 2.0" U.S. Department of Agriculture, Agricultural Research
Service, U.S. Salinity Laboratory, Riverside, CA.
SoilCover. (2000). "SoilCover User's Manual" Unsaturated Soils Group, Department of Civil
Engineering, University of Saskatchewan, Saskatoon, Canada.
Thiel, R.S. and Steward, M.G. (1993). "Geosynthetic Landfill Cover Design Methodology and
Construction Experience in the Pacific Northwest," Geosynthetics '93 Conference Proceedings,
Industrial Fabrics Association International, St. Paul, MN, Vol. 3, pp. 1131-1144.
Thompson, F.L. and Tyler, S.W. (1984). "Comparison of Two Groundwater Flow Models -
UNSAT1D and HELP and Their Application to Covered Fly Ash Disposal Sites" EPRI CS-3695,
Electric Power Research Institute, Palo Alto, CA.
Thornthwaite, C.W. and Mather, J.R. (1955). "The Water Balance," Publications in
Climatology, Drexel Institute of Technology, Laboratory of Technology, Philadelphia, PA, Vol.
8, No. 1, 104 p.
Thornthwaite, C.W. and Mather, J.R. (1957). "Instructions and Tables for Computing Potential
Evapotranspiration and the Water Balance," Publications in Climatology, Drexel Institute of
Technology, Laboratory of Technology, Philadelphia, PA, Vol. 10, No. 3, pp. 185-311.
DRAFT - DO NOT CITE OR QUOTE
A-24
-------
Tratch, D. (1994). "Moisture Uptake Within the Root Zone" M.S. Thesis, Department of Civil
Engineering, University of Saskatchewan, Saskatoon, Saskatchewan, Canada.
Udoh, F.D. (1991). Minimization of Infiltration into Mining Stockpiles Using Low Permeability
Covers, Dissertation Proposal, Department of Materials Science and Engineering, Mining
Engineering Program, University of Wisconsin, Madison, WI.
Ward, A.L. and Gee, G.W. (1997). "Performance of a Field-Scale Surface Barrier," Journal of
Environmental Quality., Vol. 26, No. 3, pp. 694-705.
Waugh, W.J., Thiede, M.E., Bates, D.J., Cadwell, L.L., Gee, G.W., and Kemp, C.J. (1994).
"Plant Cover and Water Balance in Gravel Admixtures at an Arid Waste-Burial Site," Journal of
Environmental Quality., Vol. 23, No. 4, pp. 676-684.
Webb, S.W., McCord, J.T., and Dwyer, S.F. (1997). "Prediction of Tilted Capillary Barrier
Performance," Proceedings, International Containment Technology Conference., St. Petersburg,
FL, pp. 296-302.
Wilson, G.W. (1990). "Soil Evaporative Fluxes for Geotechnical Engineering Problems" Ph.D.
Dissertation, Department of Civil Engineering, University of Saskatchewan, Saskatoon, Canada.
Zornberg, J.G. and Caldwell, J.A. (1998). "Design of Monocovers for Landfills in Arid
Locations," Proceedings, Third International Conference on Environmental Geotechnics,
ISSMFE, Lisbon, Portugal.
Chapter 5
Augenstein, D., Yazdani, R., Moore, R., and Dahl, K. R. (1997). "Yolo County Controlled
Landfill Demonstration Project", Proceedings from SI
Symposium, Monterey, California, March , pp. 43-87.
Landfill Demonstration Project", Proceedings from SWANA's 20th Annual Landfill Gas
Berglund, C. (1998). "Environmental Hydraulics of Sanitary Landfills", Journal of Hydraulic
Research, Vol. 36, No. 1, pp. 101-116.
Bogner, J. and Scott, P. (1997). "Landfill CH4 Emissions: Guidance for Field Measurements",
Papers from the 1992-1994IEA Landfill Gas Activity, Anders Lagerkvist (Ed.), The Landfill
Group, Lule University of Technology, January, pp. 29-46.
Bogner, I.E., Spokas, K.A., and Burton, E.A. (1997). "Kinetics of Methane Oxidation in a
Landfill Cover Soil: Temporal Variations, a Whole-Landfill Oxidation Experiment, and
Modeling of Net CH4 Emissions".
Bonaparte, R., Daniel, D.E., and Koerner, R.M., (2002). "Assessment and Recommendations for
Optimal Performance of Waste Containment Systems," U.S. Environmental Protection Agency,
National Risk Management Research Laboratory, Cincinnati, OH (in press).
EPA (1991) "Design and Construction of RCRA/CERCLA Final Covers" Seminar Publication,
EPA/625/4-91/025, U.S. Environmental Protection Agency, Office of Research and
DRAFT - DO NOT CITE OR QUOTE
A-25
-------
Development, Washington, D.C., 145 p.
EPA (1992). "Organic Air Emissions from Waste Management Facilities," Seminar Publication,
EPA/625/R-92/003, U.S. Environmental Protection Agency, Office of Research and
Development, Washington, D.C., 115 p.
EPA. (1997a). "Compilation of Air Pollutant Emission Factors, AP-42, Supplement C", 5th
edition, EPA Office of Air Quality Planning and Standards, Research Triangle Park, North
Carolina.
EPA(1997b). "Compilation of Air Pollutant Emission Factors, AP-42, Volume 1: Stationary
Point and Area Sources", 5th edition, EPA Office of Air Quality Planning and Standards,
Research Triangle Park, North Carolina.
EPA, Landfill Gas Emissions Model (LandGEM) Version 3.02 User's Guide, Prepared for EPA's
National Risk Management Research Laboratory, EPA-600/R-05/047, April 2005.
Hashmonay, R.A., and M.G. Yost (1999). Innovative approach for estimating fugitive gaseous
fluxes using computed tomography and remote optical sensing techniques, J. Air Waste Manage.
Assoc., 49, 966-972, 1999.
Hashmonay, R.A., D.F. Natschke, K.Wagoner, D.B. Harris, E.L.Thompson, and M.G. Yost
(2001). Field evaluation of a method for estimating gaseous fluxes from area sources using open-
path Fourier transform infrared, Environ. Sci. TechnoL, 35, 2309-2313, 2001.
Hashmonay, R.A., K. Wagoner, D.F. Natschke, D.B. Harris, and E.L. Thompson (2002). Radial
computed tomography of air contaminants using optical remote sensing, presented June 23-27,
2002 at the AWMA 95th Annual Conference and Exhibition, Baltimore, MD.
Hashmonay, R.A., M.G. Yost, and C. Wu (1999). Computed tomography of air pollutants using
radial scanning path-integrated optical remote sensing, Atmos. Environ., 33, 267-274, 1999.
Hashmonay, R.A., M.G. Yost, D.B. Harris, and E.L. Thompson (1998). Simulation study for
gaseous fluxes from an area source using computed tomography and optical remote sensing,
presented at SPIE Conference on Environmental Monitoring and Remediation Technologies,
Boston, MA, Nov., 1998, in SPIE Vol. 3534, 405-410.
Huitric, R. and Soni, R. (1997). "Making the Most of Landfill Gas Projection Rates",
Proceedings from SWANA's 20th Annual Landfill Gas Symposium, Monterey, California, March ,
pp. 117-145.
Knightley, D., Nedwell, D.B., and Cooper, M. (1995). "Capacity for Methane Oxidation in
Landfill Cover Soils Measured in Laboratory-Scale Soil Microcosms", Applied and
Environmental Microbiology, American Society for Microbiology, Vol. 61, pp. 592-601.
Kreith, F. (1994). "Handbook of Solid Waste Management", McGraw Hill.
DRAFT - DO NOT CITE OR QUOTE
A-26
-------
McBean, E.A., Rovers, F.A., and Farquhar, GJ. (1995). "Solid Waste Landfill Engineering and
Design", Prentice Hall.
Modrak, M. T.; Hashmonay, R. A.; Kagann, R. (2004). Measurement of Fugitive Emissions at a
Region I Landfill, EPA-600/R-04-001; U.S. Environmental Protection Agency, Research and
Development, Work Assignment No. 4-003, January 2004.
Modrak, M. T.; Hashmonay, R. A.; Varma R.; Kagann, R. (2005a). Evaluation of Fugitive
Emissions at a Brownfield Landfill in Ft. Collins, Colorado Using Ground-Based Optical Remote
Sensing Technology, EPA-600/R-05/042; U.S. Environmental Protection Agency, Research and
Development, Work Assignment No. 0-025, March 2005.
Modrak, M. T.; Hashmonay, R. A.; Varma R.; Kagann, R. (2005b). Evaluation of Fugitive
Emissions at a Brownfield Landfill in Colorado Springs, Colorado Using Ground-Based Optical
Remote Sensing Technology, EPA-600/R-05/041; U.S. Environmental Protection Agency,
Research and Development, Work Assignment No. 0-025, March 2005.
Murphy, M. (1998). "Landfill Gas Issues at Arid Landfills", presentation at SWANA Arid
Climate Landfill Symposium, Santa Fe, New Mexico, April.
Pierce, J.L., Beizer, M.B., Nicotri, P.J., and LaFountain, L. (1998). "Design and
Implementation of a Landfill Gas Control System at the Operating Industries Landfill Superfund
Site", Proceedings from SWANA 's 21st Annual Landfill Gas Symposium, SWANA, Austin, TX,
March 1998, pp. 41-48.
Repa, E.W. (1994). "Federal Regulation", Chapter 4 in Handbook of Solid Waste Management,
Ed. F. Kreith.
Tchobanoglous, G., Theisen, H., and Vigil, S. (1993). "IntegratedSolid Waste Management,
Engineering Principles and Management Issues", McGraw-Hill, Inc., New York, 978 p.
Thoma, E. D., R. C. Shores, E. L. Thompson, D. B. Harris, S. A. Thorneloe, R. M.
Varma, R. A. Hashmonay, M. T. Modrak, D. F. Natschke, and H. A. Gamble. (2005). Open Path
Tunable Diode Laser Absorption Spectroscopy for Acquisition of Fugitive Emission
Flux Data; Journal of Air and Waste Management Association. (55), 658-668 (2005).
Thornloe, S.A., A. Reisdorph, M. Laur, R. Pelt, R.L. Bass, and C. Burklin, (1999). The U.S.
Environmental Protection Agency's Landfill Gas Emissions Model (LandGEM), Sardinia 99,
Seventh International Waste Management and Landfill Symposium, Published in Proceedings,
Vol. IV, pages 11-18, October 4-8, 1999.
Whalen, S.C., Reeburgh, W.S., Sandbeck, K.A. (1990). "Rapid Methane Oxidation in a Landfill
Cover Soil", Applied and Environmental Microbiology, American Society for Microbiology,
November, pp. 3405-3411.
Wu, C., M.G. Yost, R.A. Hashmonay, and D.Y. Park, Experimental evaluation of a radial beam
geometry for mapping air pollutants using optical remote sensing and computed tomography,
Atmos. Environ., 33, 4709-4716, 1999.
DRAFT - DO NOT CITE OR QUOTE
A-27
-------
Chapter 6
Achilleous, E. (1988). "PC STABL5M, User Manual" Informational Report, School of Civil
Engineering, Purdue University, West Lafayette, IN, 132 p.
Ajaz, A. and Parry, R.H.G. (1975a). "Stress-Strain Behavior of Two Compacted Clays in
Tension and Compression," Geotechnique, Vol. 25, No. 3, pp. 495-512.
Ajaz, A. and Parry, R.H.G. (1975b). "Analysis of Bending Stresses in Soil Beams,"
Geotechnique, Vol. 25, No. 3, pp. 586-591.
Ajaz, A. and Parry, R.H.G. (1976). "Bending Test for Compacted Clays," Journal of the
Geotechnical Engineering Division, Vol. 102, No. 9, pp. 929-943.
Anderson, D.G. and Kavazanjian, E., Jr. (1995). "Performance of Landfills Under Seismic
Loading," Proceedings, Third International Conference on Recent Advances in Geotechnical
Earthquake Engineering and Soil Dynamics, St. Louis, MO, Vol. 3, pp. 227-306.
Augello, A.J., Matasovic, N., Bray, J.D., Kavazanjian, E., Jr., and Seed, R.B. (1995).
"Evaluation of Solid Waste Landfill Performance During the Northridge Earthquake,"
Earthquake Design and Performance of Solid Waste Landfills., Geotechnical Special Publication
No. 54, M.K. Yegian and W.D.L. Finn (eds.), ASCE, New York, NY, pp. 17-50.
Augello, A.J., Bray, J.D., Abrahamson, N.A., and Seed, R.B. (1998). "Dynamic Properties of
Solid Waste Based on Back-Analysis of Oil Landfill," Journal of Geotechnical and
GeoenvironmentalEngineering, Vol. 124, No. 3, pp. 211-222.
Bardet, J.P. (1997). "Experimental Soil Mechanics," Prentice Hall, Inc., Upper Saddle River,
NJ, 583 p.
Battelle (1993). "Technical Resource Document: Solidification/Stabilization and Its Application
to Waste Materials," EPA/530/R-93/012, U.S. Environmental Protection Agency, Risk
Reduction Engineering Laboratory, Washington, D.C.
Bemben, S.M. and Schulze, D.A. (1995). "The Influence of Testing Procedures on Clay-
Geomembrane Shear Strength Measurements," Geosynthetics '95 Conference Proceedings,
Industrial Fabrics Association International, St. Paul, MN, pp. 1043-1056.
Bishop, A.W. (1955). "The Use of the Slip Circle in the Stability Analysis of Slopes,"
Geotechnique, Vol. 5, No. 1, pp. 7-17.
Bjarngard, A. and Edgers, L. (1990). "Settlements of Municipal Solid Waste Landfills,"
Proceedings, 13th Annual Madison Waste Conference, Madison, WI.
Bodine, D.G. and Trevino, F.M. (1996). "RCRA Closure of Refinery Sludge Basin Using In-
Situ Solidification and Containment," Proceedings, Fourth Great Lakes Geotechnical/
Geoenvironmental Conference, Chicago, IL, pp. 49-74.
DRAFT - DO NOT CITE OR QUOTE
A-28
-------
Bonaparte, R. and Berg, R.R. (1987). "Long-Term Allowable Tension for Geosynthetic
Reinforcement," Geosynthetics '87 Conference Proceedings, Industrial Fabrics Association
International, St. Paul, MN, Vol. 1, pp. 108-120.
Bonaparte, R. and Christopher, B.R. (1987). "Design and Construction of Reinforced
Embankments Over Weak Foundations," Transportation Research Record 7753, Transportation
Research Board, Washington, D.C., pp. 26-39.
Bonaparte, R., Williams, N., and Giroud, J.P. (1985). "Innovative Leachate Collection Systems
for Hazardous Waste Containment Facilities," Proceedings of Geotechnical Fabrics Conference
'85, Cincinnati, OH, pp. 9-34.
Bonaparte, R., Othman, M.A., Rad, N.S., Swan, R.H., and Vander Linde, D.L. (1996).
"Evaluation of Various Aspects of GCL Performance," Appendix F in Report of 1995 Workshop
on Geosynthetic Clay Liners, D.E. Daniel and H.B. Scranton (eds.), EPA/600/R-96/149, U.S.
Environmental Protection Agency, National Risk Management Research Laboratory, Cincinnati,
OH, pp. F1-F34.
Bourdeau, P.L., Ludlow, S.J., and Simpson, B.E. (1993). "Stability of Soil-Covered
Geosynthetic-Lined Slopes: A Parametric Study," Geosynthetics '93 Conference Proceedings,
Industrial Fabrics Association International, St. Paul, MN, Vol. 2, pp. 1511-1521.
Bowders, J.J., Daniel, D.E., Wellington, J., and Houssidas, V. (1997a). "Managing Desiccation
Cracking in Compacted Clay Liners Beneath Geomembranes," Geosynthetics '97 Conference
Proceedings, Industrial Fabrics Association International, St. Paul, MN, Vol. 1, pp. 527-540.
Bowders, J.J., Tan, J.P., Daniel, D.E. (1997b). "Expanded Clay and Shale Aggregates for
Leachate Collection Systems," Journal of Geotechnical and Geoenvironmental Engineering,
Vol. 123, No. 11, pp. 1030-1034.
Bray, J.D., Augello, A.J., Leonards, G.A., Repetto, P.C., and Byrne, R.J. (1995). "Seismic
Stability Procedures for Solid Waste Landfills," Journal of Geotechnical Engineering, Vol. 121,
No. 2, pp. 139-151.
Bray, J.D. and Rathje, E.M. (1998). "Earthquake-Induced Displacements of Solid-Waste
Landfills," Journal of Geotechnical and Geoenvironmental Engineering, Vol. 124, No. 3, pp.
242-253.
Breitenbach, AJ. and Swan, R.H. (1999). "Influence of High Load Deformations on
Geomembrane Liner Interface Strengths," Geosynthetics '99 Conference Proceedings, Industrial
Fabrics Association International, St. Paul, MN, Vol. 1, pp. 517-529.
Burlingame, MJ. (1985). "Construction of a Highway on a Sanitary Landfill and Its Long-Term
Performance," Transportation Research Record 1031, Transportation Research Board,
Washington, D.C., pp. 34-40.
DRAFT - DO NOT CITE OR QUOTE
A-29
-------
Byrne, RJ. (1994). "Design Issues with Strain-Softening Interfaces in Landfill Liners,"
Proceedings, Waste Tech '94 Conference, National Solid Waste Management Association,
Charleston, SC, 26 p.
Carey, P.J., Koragappa, N., and Gurda, JJ. (1993). "Renewal of Old Sites, Overliner System
Design, A Case Study of the Brookhaven Landfill, Long Island, New York," Proceedings, Waste
Tech '93 Conference, National Solid Waste Management Association, Marina Del Rey, CA, 13
P-
Cargill, K.W. and Olen K.L. (1997). "An Innovative Final Closure System," Proceedings, 2nd
Annual Landfill Symposium., Solid Waste Association of North America, Sacramento, CA, pp.
65-79.
Cargill, K.W. and Olen K.L. (1998). "Landfill Closure Using Reinforced Soil Slopes,"
Proceedings, Sixth International Conference on Geosynthetics, Industrial Fabrics Association
International, St. Paul, MN, Vol. 1, pp. 481-486.
Chandhari, A.P. and Char, A.N.R. (1985). "Flexural Behavior of Reinforced Beams," Journal of
GeotechnicalEngineering, Vol. Ill, No. 11, pp. 1328-1333.
Chaney, R.C., Thiel, R., Richardson, G.N., and Cadwallader, M. (1997). "Cyclic Response of
Clay-Geomembrane Interfaces and Their Impact on the Seismic Response of Lined Landfills,"
Geosynthetics '97 Conference Proceedings, Industrial Fabrics Association International, St. Paul,
MN, Vol. 2, pp 977-987.
Daniel, D.E. and Scranton, H.B. (1996). "Report of 1995 Workshop on Geosynthetic Clay
Liners" EPA/600/R-96/149, U.S. Environmental Protection Agency, National Risk Management
Research Laboratory, Cincinnati, OH, 96 p.
Daniel, D.E., Shan, H.-Y., and Anderson, J.D. (1993). "Effects of Partial Wetting on the
Performance of the Bentonite Component of a Geosynthetic Clay Liner," Geosynthetics '93
Conference Proceedings., Industrial Fabrics Association International, St. Paul, MN, Vol. 3, pp.
1483-1496.
Daniel, D.E., Koerner, R.M., Bonaparte, R., Landreth, R.E., Carson, D.A., and Scranton, H.B.
(1998). "Slope Stability of Geosynthetic Clay Liner Test Plots," Journal of Geotechnical and
GeoenvironmentalEngineering, Vol. 124, No. 7, pp. 628-637.
De, A. and Zimme, T.F. (1998). "Frictional Behavior of Landfill Liner Interfaces with Geonets,"
Proceedings, Sixth International Conference on Geosynthetics, Industrial Fabrics Association
International, St. Paul, MN, Vol. 1, pp. 443-446.
Dove, I.E., Frost, J.D., Han, J., and Bachus, R.C. (1997). "The Influence of Geomembrane
Surface Roughness on Interface Strength," Geosynthetics '97 Conference Proceedings, Industrial
Fabrics Association International, St. Paul, MN, pp. 863-876.
DRAFT - DO NOT CITE OR QUOTE
A-30
-------
Druschel, SJ. and Underwood, E.R. (1993). "Design of Lining and Cover System Sideslopes,"
Geosynthetics '93 Conference Proceedings, Industrial Fabrics Association International, St. Paul,
MN, Vol. 3, pp. 1341-1355.
Duncan, J.M. (1992). "State-of-the-Art: Static Stability and Deformation Analysis," Stability
and Performance of Slopes and Embankments - II, Geotechnical Special Publication No. 31,
R.B. Seed and R.W. Boulanger (eds.), ASCE, New York, NY, Vol. 1, pp. 222-266.
Edgers, L., Noble, J. 1, and Williams, E. (1992). "A Biologic Model for Long-Term Settlement
in Landfills," Proceedings, Conference on Environmental Technology, A.A. Balkama,
Rotterdam, pp. 177-184.
Edil, T.B., Ranguette, V.J., and Wuellner, W.W. (1990). "Settlement of Municipal Refuse,"
Geotechnics of Waste Landfills - Theory and Practice, STP 1070, A. Landva and E.D. Knowles
(eds.), American Society for Testing and Materials, Philadelphia, PA, pp. 225-239.
Eid, H.T. and Stark, T.D. (1997). "Shear Behavior of an Unreinforced Geosynthetic Clay
Liner," Geosynthetics International, Vol. 4, No. 6, pp. 645-659.
Elias, V. and Christopher, B.R. (1996). "Mechanically Stabilized Earth Walls and Reinforced
Soil Slopes," Publication No. FHWA-SA-96-072, Federal Highway Administration, Washington,
D.C.
EPA (1988). "Guide to Technical Resources for the Design of Land Disposal Facilities"
EPA/625/6-88/018, U.S. Environmental Protection Agency, Risk Reduction Engineering
Laboratory, Cincinnati, OH, 63 p.
EPA (1991). "Design and Construction of RCRA/CERCLA Final Covers" Seminar Publication,
EPA/625/4-91/025, U.S. Environmental Protection Agency, Office of Research and
Development, Washington, D.C., 145 p.
Fassett, J.B., Leonards, G.A., and Repetto, P.C. (1994). "Geotechnical Properties of Municipal
Solid Wastes and Their Use in Landfill Design," Proceedings, Waste Tech '94 Conference,
National Solid Waste Management Association, Charleston, SC, 32 p.
Fox, P.J., Rowland, M.G., and Scheithe, J.R. (1998). "Internal Shear Strength of Three
Geosynthetic Clay Liners, " Journal of Geotechnical and Geoenvironmental Engineering, Vol.
124, No. 10, pp. 933-944.
Frankel, F., Mueller, C., Bernhard, T., Perkins, D., Leyendecker, E.V., Dickman, N., Hanson, S.,
and Hopper, M. (1996). Open-File Report 96-532, "National Seismic Hazard Maps:
Documentation June 1996," U.S. Geological Survey, Denver, CO.
Gaind, KJ. and Char, A.N.R. (1983). "Reinforced Soil Beams," Journal of Geotechnical
Engineering, Vol. 109, No. 7, pp. 977-982.
Gibson, R.E. and Lo, K.Y. (1961). "A Theory of Soils Exhibiting Secondary Compression,"
ActaPolytechnica Scandinavica, C, 10, 296, pp. 1-15.
DRAFT - DO NOT CITE OR QUOTE
A-31
-------
Gilbert, P.A. and Murphy, W.L. (1987). "Prediction/Mitigation of Subsidence Damage to
Hazardous Waste Landfill Covers" EPA/600/12/87-025, U.S. Environmental Protection
Agency, Hazardous Waste Engineering Research Laboratory, 81 p.
Gilbert, R.B., Liu, C.N., Wright, S.G, and Trautwein, S.J. (1995). "A Double Shear Test
Method for Measuring Interface Strength," Geosynthetics '95 Conference Proceedings., Industrial
Fabrics Association International, St. Paul, MN, Vol. 3, pp. 1017-1029.
Gilbert, R.B., Fernandez, F., and Horsfield, D.W. (1996). "Shear Strength of Reinforced
Geosynthetic Clay Liner," Journal of'GeotechnicalEngineering, Vol. 122, No. 4, pp. 259-266.
Giroud, J.P. and Beech, J.F. (1989). "Stability of Soil Layers on Geosynthetic Lining Systems,"
Geosynthetics '89 Conference Proceedings, Industrial Fabrics Association International, St.
Paul, MN, Vol. 1, pp. 35-46.
Giroud, J.P., Bonaparte, R., Beech, J.F., and Gross, B.A. (1990). "Design of Soil Layer-
Geosynthetic System Overlapping Voids," Geotextiles andGeomembranes, Vol. 9, No. 1, pp. 11-
50.
Giroud, J.P., Bachus, R.C., and Bonaparte, R. (1995a). "Influence of Water on the Stability of
Geosynthetic-Soil Layered Systems on Slopes," Geosynthetics International, Vol. 2, No. 6, pp.
1149-1180.
Giroud, J.P., Williams, N.D., Pelte, T., and Beech, J.F. (1995b). "Stability of Geosynthetic-Soil
Layered Systems on Slopes," Geosynthetics International, Vol. 2, No. 6, pp. 1115-1148.
Graves, S.M., Luettich, S.L., Duncan, P., and Kern, T.J. (1998). "Case Study - Slope
Stabilization and Flood Protection Remedy of the 50-Acre Warner Hill Landfill," Proceedings,
Waste Tech 98 Conference, San Antonio, TX, 25 p.
Gregory, G.H. and Chill, D.S. (1998). "Stabilization of Earth Slopes with Fiber Reinforcement,"
Proceedings, Sixth International Conference on Geosynthetics, Industrial Fabrics Association
International, St. Paul, MN, pp. 1073-1078.
Gross, B.A., Bonaparte, R., and Giroud, J.P. (2002). "Waste Containment Systems: Problems
and Lessons Learned," Appendix F in Assessment and Recommendations for Optimal
Performance of Waste Containment Systems, R. Bonaparte, D.E. Daniel, and R.M. Koerner, U.S.
Environmental Protection Agency, National Risk Management Research Laboratory, Cincinnati,
OH (in press).
Harder, L.S., Jr. (1991). "Performance of Earth Dams During the Loma Prieta Earthquake,"
Proceedings, Second International Conference on Recent Advances in Geotechnical Earthquake
Engineering and Soil Dynamics, St. Louis, MO, pp. 11-15.
Holtz, R.D., Christopher, B.R., and Berg, R.R. (1995). "Geosynthetic Design and Construction
Guidance," Publication No. FHWA HI-95-038, Federal Highway Administration, Washington,
D.C., 396 p.
DRAFT - DO NOT CITE OR QUOTE
A-32
-------
Holtz, R.D. and Kovacs, W.D. (1981). "AnIntroduction to GeotechnicalEngineering" Prentice
Hall, Englewood Cliffs, NJ, 733 p.
Houston, S.L., Houston, W.N., and Padilla, J.M. (1987). "Microcomputer-Aided Evaluation of
Earthquake-Induced Permanent Slope Displacements," Microcomputers in Civil Engineering,
Vol. 2, pp. 202-222.
Hudson, M., Idriss, I.M., and Beikae, M. (1994). "QUAD4M: A Computer Program to Evaluate
the Seismic Response of Soil Structures Using Finite Element Procedures and Incorporating a
Compliant Base" Center for Geotechnical Modeling, Department of Civil and Environmental
Engineering, University of California, Davis, CA.
Humphrey, D.N. and Rowe, R.K. (1991). "Design of Reinforced Embankments - Recent
Developments in the State-of-the-Art," Geotechnical Engineering Congress 1991, Geotechnical
Special Publication No. 27, F.G. McLean, D.A. Campbell, and D.W. Harris (eds.), ASCE, New
York, NY, Vol. II, pp. 1006-1020.
Hynes, M.E. and Franklin, A.G. (1984). "Rationalizing the Seismic Coefficient Method"
Miscellaneous Paper GL-84-13, U.S. Army Engineer Waterways Experiment Station, Vicksburg,
MS.
Idriss, I.M. and Sun, J.I. (1992). "SHAKE91 -A Computer Program for Conducting Equivalent
Linear Seismic Response Analyses of Horizontally Layered Soil Deposits" User's Manual,
Center for Geotechnical Modeling, Department of Civil and Environmental Engineering,
University of California, Davis, CA, 13 p. (plus appendices).
Idriss, I.M. (1990). "Response of Soft Soil Sites During Earthquakes," Proceedings, Symposium
to Honor Professor H.B. Seed, Bitech Publishers, Vancouver, B.C., Vol. 2, pp 273-290.
Idriss, I.M., Fiegel, G., Hudson, M.B., Mundy, P.K., and Herzig, R. (1995). "Seismic Response
of the Operating Industries Landfill," Earthquake Design and Performance of Solid Waste
Landfills, Geotechnical Special Publication No. 54, M.K. Yegian and W.D.L. Finn (eds.), ASCE,
New York, NY, pp. 83-118.
Jang, D.J. and Montero, C. (1993). "Design of Liner Systems Under Vertical Expansions: An
Alternative to Geogrids," Geosynthetics '93 Conference Proceedings, Industrial Fabrics
Association International, St. Paul, MN, Vol. 3, pp. 1497-1510.
Jessberger, H.L. and Stone, K. (1991). "Subsidence Effects on Clay Barriers," Geotechnique,
Vol. 41, No. 2, pp. 185-194.
Kavazanjian, E., Jr. (1998). "Current Issues in Seismic Design of Geosynthetic Cover Systems,"
Proceedings, Sixth International Conference on Geosynthetics, Industrial Fabrics Association
International, St. Paul, MN, Vol. 1, pp. 219-226.
Kavazanjian, E., Jr. and Matasovic, N. (1995). "Seismic Analysis of Solid Waste Landfills,"
Geoenvironment 2000, Geotechnical Special Publication No. 46, D.E. Daniel and Y.B. Acar
(eds.), ASCE, New York, NY, Vol. 2, pp. 1066-1080.
DRAFT - DO NOT CITE OR QUOTE
A-33
-------
Kavazanjian, E., Jr., Hushmand, B., and Martin, G.R. (1991). "Frictional Base Isolation Using a
Layered Soil-Synthetic Liner System," Proceedings, Third U.S. Conference on Lifeline
Earthquake Engineering, Monograph No. 4, Los Angeles, CA, pp. 1140-1151.
Kavazanjian, E., Jr., Snow, M.S., Matasovic, N., Poran, C., and Satoh, T. (1994). "Non-Intrusive
Rayleigh Wave Investigations at Solid Waste Landfills," Proceedings, 1st International
Conference on Environmental Geotechnics, Edmonton, Alberta, pp. 707-712.
Kavazanjian, E., Jr., Bonaparte, R., Johnson, G.W., Martin, G.R., and Matasovic, N. (1995a).
"Hazard Analysis for a Large Regional Landfill," Earthquake Design and Performance of Solid
Waste Landfill, Geotechnical Special Publication No. 54, M.K. Yegian and W.D.L. Finn (eds.),
ASCE, New York, NY, pp. 119-141.
Kavazanjian, E., Jr., Matasovic, N., Bonaparte, R., and Schmertmann, G.R. (1995b).
"Evaluation of MSW Properties for Seismic Analysis," Geoenvironment 2000, Geotechnical
Special Publication No. 46, D.E. Daniel and Y.B. Acar (eds.), ASCE, New York, NY, Vol. 2, pp.
1126-1141.
Kavazanjian, E., Jr., Matasovic, N., Stokoe, K.H., and Bray, J.D. (1996). "In Situ Shear Wave
Velocity of Solid Waste from Surface Wave Measurements," Proceedings, 2nd International
Conference on Environmental Geotechnics, Osaka, Japan, Vol. 1, pp. 97-102.
Kavazanjian, E., Jr., Matasovic, N., Hadj-Hamou, T., and Sabatini, P.J. (1997). "Geotechnical
Earthquake Engineering for Highways, Vol. 1 - Design Principles" Geotechnical Engineering
Circular #3, Publication No. FHWA-SA-97-076, Federal Highway Administration, Washington,
D.C., 186 p.
Koerner, R.M. (1998). "Designing With Geosynthetics," 4th Edition, Prentice Hall, Englewood
Cliffs, NJ, 761 p.
Koerner, R.M. and Daniel, D.E. (1997). "Final Covers for Solid Waste Landfills and Abandoned
Dumps," ASCE Press, Reston, VA, 256 p.
Koerner, R.M. and Hwu, B.L. (1991). "Stability and Tension Considerations Regarding Cover
Soils on Geomembrane Lined Slopes," Geotextiles andGeomembranes, Vol. 10, No. 4, pp. 335-
355.
Koerner, R.M. and Soong, T.Y. (1998). "Analysis and Design of Veneer Cover Soils,"
Proceedings, Sixth International Conference on Geosynthetics, Industrial Fabrics Association
International, St. Paul, MN, Vol. 1, pp. 1-26.
Konig, D., Kockel, R., and Jessberger, H. L. (1996). "Zur Beutellung der Standsicherhert und
zur Prognose der Setzungen von Mischabfalldeponien," Proc. 12th Numberg Deponieseminar,
Vol. 75, Eigenverlag LGA, Nurnberg, Germany, pp. 95-117.
Kramer, S.L. (1996). "GeotechnicalEarthquake Engineering," Prentice-Hall, Inc., Englewood
Cliffs, NJ, 653 p.
DRAFT - DO NOT CITE OR QUOTE
A-34
-------
Krinitzsky, E.L., Gould, J.P., and Edinger, P.H. (1993). "Fundamentals of Earthquake-Resistant
Construction" John Wiley & Sons, Inc., New York, NY, 299 p.
Ladd, C.C. (1991). "Stability Evaluation During Staged Construction," Journal of Geotechnical
Engineering, Vol. 117, No. 4, pp. 537-615.
LaGatta, M.D., Boardman, B.T., Cooley, B.H., and Daniel, D.E. (1997). "Geosynthetic Clay
Liners Subjected to Differential Settlement," Journal of Geotechnical and Geoenvironmental
Engineering, Vol. 123, No. 5, pp. 402-410.
Lai, J., Daniel, D.E., and Wright, S.G. (1998). "Effects of Cyclic Loading on Internal Shear
Strength of Unreinforced Geosynthetic Clay Liner," Journal of Geotechnical and
Geoenvironmental Engineering, Vol. 124, No. 1, pp. 45-52.
Lambe, T.W. (1951). "Soil Testing for Engineers;' John Wiley & Sons, Inc., New York, NY,
165 p.
Lambe, T.W. and Whitman, R.V. (1969). "SoilMechanics;' John Wiley & Sons, Inc., New
York, NY, 553 p.
Landva, A.O. and Clark, J.I. (1990). "Geotechnics of Waste Fill," Geotechnics of Waste
Fills-Theory and Practice, STP 1070, A. Landva and G.D. Knowles (eds.), American Society for
Testing and Materials, Philadelphia, PA, pp. 86-106.
Leonards, G.A. andNarain, J. (1963). "Flexibility of Clay and Cracking of Earth Dams,"
Journal of the Soil Mechanics and Foundations Division, Vol. 89, No. 2, pp. 47-98.
Li, M. and Gilbert, R.B. (1999). "Shear Strength of Textured Geomembrane andNonwoven
Geotextile Interfaces," Geosynthetics '99 Conference Proceedings, Industrial Fabrics
Association International, St. Paul, MN, Vol. 1.
Liu, C.N., Gilbert, R.B., Thiel, R.S., and Wright, S.G. (1997). "What is an Appropriate Factor of
Safety for Landfill Cover Slopes," Geosynthetics '97 Conference Proceedings, Industrial Fabrics
Association International, St. Paul, MN, Vol. 1, pp. 481-496.
Long, J.H., Daly, J.J., and Gilbert, R.B. (1993). "Structural Integrity of Geosynthetic Lining and
Cover Systems for Solid Waste Landfills," Project No. OSWR 06-005, Department of Civil
Engineering, University of Illinois, Champaign-Urbana, IL.
Long, J.H., Gilbert, R.B., and Daly, J.J. (1994). "Geosynthetic Loads in Landfill Slopes:
Displacement Compatibility," Journal of Geotechnical Engineering, Vol. 120, No. 11, pp. 2009-
2025.
Lozano, N. and Aughenbaugh, N.R. (1995). "Flexibility of Fine-Grained Soils,"
Geoenvironment 2000, Geotechnical Special Publication No. 46, D.E. Daniel and Y.B. Acar
(eds.), ASCE, New York, NY, Vol. 1, pp. 844-858.
DRAFT - DO NOT CITE OR QUOTE
A-35
-------
Marr, W.A. (2001). "Interface and Internal Shear Testing Procedures to Obtain Peak and
Residual Values," Proceedings, GRI-15 Conference on Hot Topics in Geosynthetics - II
(Peak/Residual; RECMs; Installation Concerns), Geosynthetic Research Institute, Folsom, PA,
pp. 1-28.
Martin, J.P., Koerner, R.M., and Whitty, I.E. (1984). "Experimental Friction Evaluation of
Slippage Between Geomembranes, Geotextiles, and Soils," Proceedings, International
Conference on Geomembranes, Industrial Fabrics Association International, St. Paul, MN, pp.
191-196.
Matasovic, N. (1993). "Seismic Response of Composite Horizontally-Layered Soil Deposits"
Ph.D. Dissertation, Department of Civil Engineering, University of California, Los Angeles, CA.
Matasovic, N. and Kavazanjian, E., Jr. (1998). "Cyclic Characterization of Oil Landfill Solid
Waste," Journal of Geotechnical and Geoenvironmental Engineering., Vol. 124, No. 3, pp. 197-
210.
Matasovic, N. and Vucetic, M. (1995). "Seismic Response of Soil Deposits Composed of Fully-
Saturated Clay and Sand Layers," Proceedings, 1st International Conference on Earthquake
Geotechnical Engineering., Tokyo, Japan, Vol. 1, pp. 611-616.
Matasovic, N., Kavazanjian, E., Jr., and Yan, L. (1997). "Newmark Deformation Analysis with
Degrading Yield Acceleration," Geosynthetics '97 Conference Proceedings, Industrial Fabrics
Association International, St. Paul, MN, Vol. 2, pp. 989-1000.
Matasovic, N., Kavazanjian, E., Jr., and Anderson, R.L. (1998). "Performance of Solid Waste
Landfills in Earthquakes," Earthquake Spectra, Vol. 14, No. 2, pp. 319-334.
Matasovic, N., Kavazanjian, E., Jr., Bonaparte, R., and Giroud, J.P. (2002). "Approximate
Solution for Pseudo-Static Seismic Analysis for Landfill Cover Systems," Geosynthetics
International (in press).
McKelvey, J.A. and Deutsch, W.L. (1991). "The Effect of Equipment Loading and Tapered Soil
Layers on Geosynthetic Lined Landfill Slopes," Proceedings, 14th Annual Madison Waste
Conference, University of Wisconsin, Madison, WI, pp. 395-411.
Michalski, S.R., Szwed, D.F., Kay, G.P., Knott, D.L., and Newman, F.B. (1995). "Closure of
Pickle Liquor Sludge Lagoon," Geoenvironment 2000, Geotechnical Special Publication No. 46,
D.E. Daniel and Y.B. Acar (eds.), ASCE, New York, NY, Vol. 2, pp. 1750-1764.
Morris, D.V. and Woods, C.E. (1990). "Settlement and Engineering Considerations in Landfill
Final Cover Design," Geotechnics of Waste Landfills - Theory and Practice, A. Landva and E.D.
Knowles (eds.), STP 1070, American Society for Testing and Materials, Philadelphia, PA, pp. 9-
12.
Morgenstern, N.R. and Price, V.E. (1965). "The Analysis of the Stability of General Slip
Surfaces," Geotechnique, Vol. 15, No. 1, pp. 79-93.
DRAFT - DO NOT CITE OR QUOTE
A-36
-------
Murphy, W.L. and Gilbert, P. A. (1985). "Settlement and Cover Subsidence of Hazardous Waste
Landfill" EPA/600/2-85/035, U.S. Environmental Protection Agency, Hazardous Waste
Engineering Research Laboratory, Cincinnati, OH, 110 p.
Nagasawa, T. and Umeda, Y. (1985) "Effects of the Freeze-Thaw Process on Soil Structure,"
Ground Freezing 1985, A. A. Balkama, Rotterdam, 2, pp. 219-225.
Nataraj, M.S., Magauti, R.S., and McManis, K.L. (1995). "Interface Frictional Characteristics of
Geosynthetics," Geosynthetics '95 Conference Proceedings, Industrial Fabrics Association
International, St. Paul, MN, Vol. 3, pp. 1057-1069.
NAVFAC (1983). "Soil Dynamics, Deep Stabilization, and Special Geotechnical Construction,
Design Manual 7.3" U.S. Navy Facilities Command, pp. 77-78.
Newmark, N.M. (1965). "Effects of Earthquakes on Dams and Embankments," Geotechnique,
Vol. 15, No. 2, pp. 139-160.
Othman, M.A., Benson, C.H., Chamberlain, E.J., andZimmie, T.F. (1994). "Laboratory Testing
to Evaluate Changes in Hydraulic Conductivity of Compacted Clays Caused by Freeze-Thaw:
State-of-the-Art," Hydraulic Conductivity and Waste Containment Transport in Soils, STP 1142,
D.E. Daniel and SJ. Trautwein (eds.), American Society for Testing and Materials, Philadelphia,
PA, pp. 227-254.
Othman, M.A., Bonaparte, R., Gross, B.A., and Schmertmann, G.R. (1995). "Design of MSW
Landfill Final Cover Systems," Landfill Closures...Environmental Protection and Land
Recovery, Geotechnical Special Publication No. 53, RJ. Dunn and U.P. Singh (eds.), ASCE,
New York, NY, pp. 218-257.
Pasqualini, E., Roccato, M., and Sani, D. (1993). "Shear Resistance of the Interfaces of
Composite Liners," Proceedings, Sardinia '93 Fourth International Landfill Symposium,
Cagliari, Italy, Vol. 2, pp. 1457-1471.
Poorooshasb, H.B. (1991). "Load Settlement Response of a Compacted Fill Layer Supported by
a Geosynthetic Overlying a Void," Geotextiles andGeomembranes, Vol. 10, No. 3, pp. 179-201.
Rao, S.K., Moulton, L.K., and Seals, R.K. (1977). "Settlement of Refuse Landfills,"
Geotechnical Practice Waste Disposal '87, Geotechnical Special Publication No. 13, R.D.
Woods (ed.), ASCE, New York, NY, pp. 574-598.
Reiter, L. (1990). "Earthquake Hazard Analysis - Issues and Insights," Columbia University
Press, New York, NY, 254 p.
Richardson, G.N., Kavazanjian, E., Jr., and Matasovic, N., (1995). "RCRA Subtitle D (258)
Seismic Design Guidance for Municipal Solid Waste Landfill Facilities," EPA/600/R-95/051,
U.S. Environmental Protection Agency, Office of Research and Development, Washington,
D.C., 143 p.
DRAFT - DO NOT CITE OR QUOTE
A-37
-------
Sabatini, P.J., Elias, V., Schmertmann, G.R., and Bonaparte, R. (1997). "Geotechnical
Engineering Circular No. 2, Earth Retaining Systems" Publication No. FHWA-SA-96-038,
Federal Highway Administration, Washington, D.C., 161 p.
Sabatini, P.J., Schmertmann, G.R., and Swan, R.H. (1998). "Issues in Clay/Textured
Geomembrane Interface Testing," Proceedings, Sixth International Conference on
Geomembranes, Industrial Fabrics Association International, St. Paul, MN, Vol. 1, pp. 423-426.
Sabatini, P.J., Griffin, L.M., Bonaparte, R., Espinoza, R.D., Giroud, J.P. (2001). "Reliability of
State-of-Practice for Selection of Shear Strength Parameters for Waste Containment System
Stability Analysis," Proceedings, GRI-15 Conference on Hot Topics in Geosynthetics - II
(Peak/Residual; RECMs; Installation Concerns), Geosynthetic Research Institute, pp 86-109.
Sagaseta, C. (1987). "Analysis of Undrained Soil Deformation Due to Ground Loss,"
Geotechnique, Vol. 37, No. 3, pp. 301-320.
Saxena, S.K. and Wong, Y.T. (1984). "Frictional Characteristics of a Geomembrane,"
Proceedings, International Conference on Geomembranes, Industrial Fabrics Association
International, St. Paul, MN, Vol. 1, pp. 187-190.
Schnabel, P.B., Lysmer, 1, and Seed, H.B. (1972). "SHAKE: A Computer Program for
Earthquake Response Analysis of Horizontally Layered Sites" Report No. EERC 72-12,
Earthquake Engineering Research Center, University of California, Berkeley, CA.
Seed, H.B. and Idriss, I.M. (1982). "GroundMotions and Soil Liquefaction During
Earthquakes" Monograph No. 5, Earthquake Engineering Research Institute, University of
California, Berkeley, CA, 134 p.
Seed, R.B., Mitchell, J.K., and Seed, H.B. (1988). "Slope Stability Failure Investigation:
Landfill Unit B-19, Phase I-A, Kettleman Hills, California," Report No. UCB/GT/88-01,
University of California at Berkeley, Berkeley, CA.
Seed, R.B. and Boulanger, R.W. (1991). "Smooth HDPE-Clay Liner Interface Shear Strengths:
Compaction Effects," Journal of GeotechnicalEngineering, Vol. 117, No. 4, pp. 686-693.
Seed, R.B., Dickenson, S., and Mok, C.M. (1991). "Seismic Response Analyses of Soft and
Deep Cohesive Sites: A Brief Summary of Recent Lessons," Proceedings, 1st Annual
CALTRANS Seismic Research Seminar, CALTRANS, Sacramento, CA.
Seed, R.B. and Bonaparte, R. (1992). "Seismic Analysis and Design of Lined Waste Fills:
Current Practice," Stability and Performance of Slopes and Embankments - II, Geotechnical
Special Publication No. 31, R.B. Seed and R.W. Boulanger (eds.), ASCE, New York, NY, Vol.
2, pp. 1521-1545.
Shallenberger, W.C. and Filz, G.M. (1996). "Interface Strength Determination Using a Large
Displacement Shear Box," Proceedings, 2nd International Conference on Environmental
Geotechnics, Osaka, Japan, Vol. 1, pp. 147-152.
DRAFT - DO NOT CITE OR QUOTE
A-38
-------
Sharma, H.D., Dukes, M.T., and Olsen, D.M. (1990). "Field Measurements of Dynamic Moduli
and Poisson's Ratios of Refuse and Underlying Soils at a Landfill Site," Geotechnics of Waste
Landfills - Theory and Practice, A. Landva and E.D. Knowles (eds.), STP 1070, American
Society for Testing and Materials, Philadelphia, PA, pp. 57-70.
Sharma, H.D., Hullings, D.E., and Greguras, F.D. (1997). "Interface Strength Tests and
Application to Landfill Design," Geosynthetics '97 Conference Proceedings, Industrial Fabrics
Association International, St. Paul, MN, Vol. 2, pp. 913-926.
Sharma, H.D. and Lewis, S.P. (1994). "Waste Containment Systems, Waste Stabilization, and
Landfills, Design and Evaluation" John Wiley & Sons, Inc., New York, NY, 588 p.
Sharma, S. (1994). "XSTABL, Reference Manual, Version 5," Interactive Software Designs, Inc.,
Moscow, ID.
Singh, S. and Sun, J.I. (1995). "Seismic Evaluation of Municipal Solid Waste Landfills,"
Geoenvironment 2000, Geotechnical Special Publication No. 46, D.E. Daniel and Y.B. Acar
(eds.), ASCE, New York, NY, Vol. 2, pp. 1081-1096.
Snow, M., Mansour, R., Swan, R.H., Jr., and Kavazanjian, E., Jr. (1998). "Variability of
Interface Shear Strengths," Proceedings, Sixth International Conference on Geomembranes,
Industrial Fabrics Association International, St. Paul, MN, Vol. 1, pp. 439-442.
Soong, T.-Y. and Koerner, R.M. (1997). "The Design of Drainage Systems Over
Geosynthetically Lined Slopes," GRI Report No. 19, Geosynthetic Research Institute, Drexel
University, Philadelphia, PA, 88 p.
Sowers, G.F. (1973). "Settlement of Waste Disposal Fills," Proceedings, Eighth International
Conference on Soil Mechanics and Foundation Engineering, Moscow, 207-210.
Spencer, E. (1967). "A Method of Analysis of the Stability of Embankments Assuming Parallel
Interstice Forces," Geotechnique, Vol. 17, pp. 11-26.
Spikula, D. (1996). "Subsidence Performance of Landfills: A 7-Year Review," Proceedings,
GRI-10 Conference on Field Performance of Geosynthetics and Geosynthetic Related Systems,
Geosynthetic Research Institute, Philadelphia, PA, pp. 237-244.
Stark, T.D., Williamson, T.A., and Eid, H.T. (1996). "HOPE Geomembrane/Geotextile Interface
Shear Strength," Journal of Geotechnical Engineering, Vol. 122, No. 3, pp. 197-203.
Stark, T.D. and Eid, H.T. (1996). "Shear Behavior of Reinforced Geosynthetic Clay Liners,"
Geosynthetics International, Vol. 3, No. 6, pp. 771-786.
Stark, T.D. and Poeppel, A.R. (1994). "Landfill Liner Interface Strengths from Torsional-Ring-
Shear Tests," Journal of Geotechnical Engineering, Vol. 120, No. 3, pp. 597-615.
DRAFT - DO NOT CITE OR QUOTE
A-39
-------
Stulgis, R.P., Soydemir, C., and Telgener, RJ. (1995). "Predicting Landfill Settlement,"
Geoenvironment 2000, Geotechnical Special Publication No. 46, D.E. Daniel and Y.B. Acar
(eds.), ASCE, New York, NY, Vol. 2, pp. 980-993.
Swan, R.H., Bonaparte, R., Bachus, R.C., Rivette, C.A., and Spikula, D.R. (1991). "Effect of
Soil Compaction Conditions on Geomembrane-Soil Interface Strength," Geotextiles and
Geomembranes, Vol. 10, No. 5, pp. 523-529.
Williams, N.D. and Houlihan, M.F. (1987). "Evaluation of Interface Friction Properties
Between Geosynthetics and Soils," Geosynthetics '87 Conference Proceedings, Industrial
Fabrics Association International, St. Paul, MN, Vol. 2, pp. 616-627.
Wilson-Fahmy, R.F. and Koerner, R.M. (1992). "Stability Analysis of Multi-Lined Slopes in
Landfill Applications," GRI Report No. 8, Geosynthetic Research Institute, Drexel University,
Philadelphia, PA.
Wilson-Fahmy, R.F. and Koerner, R.M. (1993). "Finite Element Analysis of Stability of Cover
Soil on Geomembrane-Lined Slopes," Geosynthetics '93 Conference Proceedings, Industrial
Fabrics Association International, St. Paul, MN, Vol. 3, pp. 1425-1438.
Wright, S.G. (1991). "UTEXAS3 -A Computer Program for Slope Stability Calculations"
Department of Civil Engineering, University of Texas, Austin, TX.
Yan, L., Matasovic, N., and Kavazanjian, E., Jr. (1996). "YSLIP_PM - A Computer Program for
Simulation of Dynamic Behavior of a Rigid Block on an Inclined Plane and Calculation of
Permanent Displacements of the Block," User's Manual, GeoSyntec Consultants, Huntington
Beach, CA, 21 p. (plus appendix).
Yegian, M.K. and Lahalf, A.M. (1992). "Dynamic Interface Shear Strength Properties of
Geomembranes and Geotextiles," Journal of Geotechnical Engineering, Vol. 118, No. 5, pp.
760-779.
Yegian, M.K., Yee, Z.Y., and Harb, J.N. (1995). "Seismic Response of Geosynthetic/Soil
Systems," Geoenvironment 2000, Geotechnical Special Publication No. 46, D.E. Daniel and
Y.B. Acar (eds.), ASCE, New York, NY, Vol. 2, pp. 1113-1125.
Yen, B.C. and Scanlon, B.S. (1975). "Sanitary Landfill Settlement Rates," Journal of the
Geotechnical Engineering Division, Vol. 101, No. 5, pp. 475-487.
Chapter 7
Augello, A.J., Matasovic, N., Bray, J.D., Kavazanjian, E., Jr., and Seed, R.B. (1995).
"Evaluation of Solid Waste Landfill Performance During the Northridge Earthquake,"
DRAFT - DO NOT CITE OR QUOTE
A-40
-------
Earthquake Design and Performance of Solid Waste Landfills., Geotechnical Special Publication
No. 54, M.K. Yegian and W.D.L. Finn (eds.), ASCE, New York, pp. 17-50.
Bass, J.M. (1986). "Avoiding Failure ofLeachate Collection and Cap Drainage Layers"
EPA/600/2-86/058, EPA Hazardous Waste Engineering Research Laboratory, Cincinnati, OH,
142 p.
Bonaparte, R., Othman, M.A., Rad, N.R., Swan, R.H., and Vander Linde, D.L. (1996).
"Evaluation of Various Aspects of GCL Performance," Appendix F in Report of 1995 Workshop
on Geosynthetic Clay Liners, D.E. Daniel and H.B. Scranton (eds.), EPA/600/R-96/149, U.S.
Environmental Protection Agency, National Risk Management Research Laboratory, Cincinnati,
OH, pp. F1-F34.
Bonaparte, R., Daniel, D.E., and Koerner, R.M., (2002). "Assessment and Recommendations for
Optimal Performance of Waste Containment Systems" U.S. Environmental Protection Agency,
National Risk Management Research Laboratory, Cincinnati, OH (in press).
Boschuk, J., Jr., (1991). "Landfill Covers An Engineering Perspective," Geotechnical Fabrics
Report, Vol. 9, No. 4, Industrial Fabrics Association International, St. Paul, MN, pp. 23-34.
Buranek, D. and Prasad, S. (1991). "Sanitary Landfill Performance During the Loma Prieta
Earthquake," Proceedings, Second International Conference on Recent Advances in
Geotechnical Earthquake Engineering and Soil Dynamics, St. Louis, MO, pp. 1655-1660.
Calabria, C.R. and Peggs, ID. (1996). "Investigation of Geomembrane Seam Failures: Landfill
Cover," Proceedings, 10th GRI Conference Field Performance of Geosynthetics and
Geosynthetic Related Systems, Geosynthetic Research Institute, Philadelphia, PA, pp. 234-257.
Carson, D.A., Bonaparte, R., Daniel, D.E., and Koerner, R.M. (1998). "Geosynthetic Clay
Liners at Field-Scale: Internal Shear Test Progress," Proceedings, Sixth International
Conference on Geosynthetics, Vol. I, Industrial Fabrics Association International, St. Paul, MN,
pp. 427-432.
Corser, P. and Cranston, M. (1991). "Observations on Long-Term Performance of Composite
Clay Liners and Covers," Proceedings, Geosynthetic Design and Performance, Vancouver
Geotechnical Society, Vancouver, BC.
Corser, P., Pellicer, J., and Cranston, M. (1992). "Observation on Long-Term Performance of
Composite Clay Liners and Covers," Geotechnical Fabrics Report, Vol. 10, No. 8, Industrial
Fabrics Association International, St. Paul, MN, pp. 6-16.
Daniel, D.E. and Gross, B.A. (1996), "Caps", Section 6 in Assessment of Barrier Containment
Technologies, R.R. Rumer and J.K. Mitchell (eds.), National Technical Information Services,
PB96-180583, Springfield, VA, pp. 119-140.
Daniel, D.E. and Koerner, R.M. (1993). "Technical Guidance Document: Quality Assurance
and Quality Control for Waste Containment Facilities," EPA/600/R-93/182, U.S. Environmental
Protection Agency, Risk Reduction Research Laboratory, Cincinnati, OH, 305 p.
DRAFT - DO NOT CITE OR QUOTE
A-41
-------
Daniel, D.E. (2002). "Cincinnati Geosynthetic Clay Liner Test Site," Appendix D in Assessment
and Recommendations for Optimal Performance of Waste Containment Systems, R. Bonaparte,
D.E. Daniel, and R.M. Koerner, U.S. Environmental Protection Agency, National Risk
Management Research Laboratory, Cincinnati, OH (in press).
Daniel, D.E., Koerner, R.M., Bonaparte, R., Landreth, R.E., Carson, D.A., and Scranton, H.B.
(1998). "Slope Stability of Fourteen Full-Scale Test Plots Containing Geosynthetic Clay
Liners," Journal of Geotechnical and GeoenvironmentalEngineering, ASCE, Vol. 124, No. 7,
pp. 628-637.
Dwyer, S.F. (1997). "Cost Comparisons of Alternative Landfill Covers," Proceedings,
International Containment Technology Conference, St. Petersburg, FL, pp. 400-406.
Dwyer, S.F. (1997). "Sunrise Mountain Landfill Final Cover Evaluation," Bureau of Land
Management Report, Las Vegas, Nevada, November 13, 1997.
Dwyer, S.F. (1998). "Alternative Landfill Covers Pass the Test," Civil Engineering, ASCE, Vol.
68, No. 9, pp 50-52.
Dwyer, S.F. (2001). "Finding a Better Cover," Civil Engineering, ASCE, Vol. 71, No. 1, pp 58-
63.
Giroud, J.P. (1994). "Quantification of Geosynthetics Behavior," Proceeding of the Fifth
International Conference on Geotextiles, Geomembranes, and Related Products, Singapore, Vol.
4, pp. 1249-1273.
Giroud, J.P. (1996). "Granular Filters and Geotextile Filters," Proceedings, Geofilters '96, J.
Lafleur and A.L. Rollin (eds.), Montreal, pp. 565-690.
Gross, B.A., Bonaparte, R., and Giroud, J.P. (2002). "Waste Containment Systems: Problems
and Lessons Learned," Appendix F in Assessment and Recommendations for Optimal
Performance of Waste Containment Systems, R. Bonaparte, D.E. Daniel, and R.M. Koerner, U.S.
Environmental Protection Agency, National Risk Management Research Laboratory, Cincinnati,
OH (in press).
Harris, J.M., Rivette, C.A., and Spradley, G.V. (1992). "Case Histories of Landfill Erosion
Protection Using Geosynthetics," Geotextiles and Geomembranes, No. 11, pp. 573-585.
Johnson, M.E., Lundy, J., Lew, M., and Ray, M.E. (1991). "Investigation of Sanitary Landfill
Slope Performance During Strong Ground Motion From the Loma Prieta Earthquake of October
17, 1989," Proceedings, Second International Conference on Recent Advances in Geotechnical
Earthquake Engineering and Soil Dynamics, St. Louis, MO, pp. 1701-1706.
Khire, M. V. (1995). "Field Hydrology and Water Balance Modeling of Earthen Final Covers
for Waste Containment," Environmental Geotechnics Report No. 95-5, Geotechnical
Engineering Program, Department of Civil & Environmental Engineering, University of
Wisconsin, Madison, WI, 166 p.
DRAFT - DO NOT CITE OR QUOTE
A-42
-------
Khire, M.V., Benson, C.H., and Bosscher, PJ. (1997). "Water Balance Modeling of Earthen
Final Covers," Journal of Geotechnical and GeoenvironmentalEngineering, Vol. 123, No. 8, pp.
744-754.
Khire, M.V., Benson, C.H., and Bosscher, PJ. (1999). "Field Data from a Capillary Barrier and
Model Predictions with UNSAT-H," Journal of Geotechnical and Geoenvironmental
Engineering, Vol. 125, No. 6, pp. 518-527.
Koerner, G.R., Koerner, R.M., and Martin, J.P. (1993). "Design of Landfill Leachate-Collection
Filters," Journal of Geotechnical Engineering, Vol. 120, No. 10, pp. 1792-1803.
Koerner, R.M., Soong, T.-Y., and Koerner, G.R. (2002). "Behavior of Waves in HOPE
Geomembranes," Appendix A in Assessment and Recommendations for Optimal Performance of
Waste Containment Systems, R. Bonaparte, D.E. Daniel, and R.M. Koerner, U.S. Environmental
Protection Agency, National Risk Management Research Laboratory, Cincinnati, OH (in press).
Lane, D.T., Benson, C.H., and Bosscher, PJ. (1992). "Hydrologic Observations and Modeling
Assessments of Landfill Covers," Environmental Geotechnics Report No. 92-10, Geotechnical
Engineering Program, Department of Civil & Environmental Engineering, University of
Wisconsin, Madison, WI, 406 p.
Maine Bureau of Remediation and Waste Management (1997). "An Assessment of Landfill
Cover System Barrier Layer Hydraulic Performance," Augusta, ME.
Matasovic, N., Kavazanjian, E., Jr., Augello, A.J., Bray, J.D., and Seed, R.B. (1995). "Solid
Waste Landfill Damage Caused by 17 January 1994 Northridge Earthquake," The Northridge,
California, Earthquake of 17 January 1994, M.C. Woods and W.R. Seiple (eds.), California
Department of Conservation, Division of Mines and Geology Special Publication 116, pp. 221-
229.
Matasovic, N. and Kavazanjian, E., Jr. (1996). "Observations of the Performance of Solid Waste
Landfills During Earthquakes," Eleventh World Conference on Earthquake Engineering,
Acapulco, Mexico, Elsevier Science Ltd., CD-ROM Paper No. 341.
Melchior, S. (1997a). "In-situ Studies on the Performance of Landfill Caps (Compacted Soil
Liners, Geomembranes, Geosynthetic Clay Liners, Capillary Barriers)," Proceedings,
International Containment Technology Conference, St. Petersburg, FL, pp. 365-373.
Melchior, S. (1997b). "In-situ Studies of the Performance of Landfill Caps (Compacted Soil
Liners, Geomembranes, Geosynthetic Clay Liners, Capillary Barriers)," Land Contamination
and Reclamation, Vol. 5, No. 3, pp. 209-216.
Melchior, S., Berger, K., Vielhaver, B., and Miehlich, G. (1994). "Multilayered Landfill Covers:
Field Data on the Water Balance and Liner Performance," In-Situ Remediation: Scientific Basis
for Current and Future Technologies, G.W. Gee andN.R. Wing (eds.), Battelle Press,
Columbus, OH, Part 1, pp. 411-425.
DRAFT - DO NOT CITE OR QUOTE
A-43
-------
Montgomery, RJ. and Parsons, LJ. (1989). '"Hie Omega Hills Final Cover Test Plot Study:
Three-Year Data Summary" presented at Annual Meeting of the National Solid Waste
Management Association, Washington, DC.
Montgomery, RJ. and Parsons, LJ. (1990). "The Omega Hills Final Cover Test Plot Study:
Fourth Year Data Summary" presented at 22nd Mid-Atlantic Industrial Waste Conference,
Drexel University, Philadelphia, PA.
Nyhan, J.W., Schofield, T.G., and Starmer, R.H. (1997). "Water Balance Study of Four Landfill
Cover Designs Varying in Slope for Semi-Arid Regions," Journal of Environmental Quality,
Vol. 26, No. 5, pp. 1385-1392.
Orr, W.R. and Finch, M.O. (1990). "Solid Waste Landfill Performance During Loma Prieta
Earthquake," Geotechnics of Waste Landfills - Theory and Practice, A. Landva and E.D.
Knowles (eds.), STP 1070, American Society for Testing and Materials, Philadelphia, PA, pp.
22-30.
Paulson, J.N. (1993). "Veneer Stability Case Histories: Design Interactions Between
Manufacturer/Consultant/Owner," Proceedings, 7th GRI Seminar-Geosynthetic Liner Systems:
Innovations, Concerns, and Design, Geosynthetic Research Institute, Philadelphia, PA, pp. 235-
241.
Soong, T.-Y. and Koerner, R.M (1997). "The Design of Drainage Systems Over
Geosynthetically Lined Slopes," GRI Report #19, Geosynthetic Research Institute, Philadelphia,
PA, 88 p.
Vander Linde, D.L., Luettich, S.M., and Bonaparte, R. (2002). "Lessons Learned from the
Failure of a Landfill Cover System," Geosynthetics: Lessons Learnedfrom Failures, J.P. Giroud,
K.L. Soderman, and G.P. Raymond (eds.). International Geosynthetics Society (in press).
Chapter 8
Anderson, I.E., Nowak, R.S., Ratzlaff, T.D., and Markham, O.D. (1993). "Managing Soil
Moisture on Waste Burial Sites in Arid Regions" Journal of Environmental Quality, Vol. 22, pp.
62-69.
Bogner, J. and Scott, P. (1997). "Landfill CH4 Emissions: Guidance for Field Measurements",
Papers from the 1992-1994IEA Landfill Gas Activity, Ed. Anders Lagerkvist, The Landfill
Group, Lule University of Technology, January, pp. 29-46.
Bonaparte, R., Daniel, D.E., and Koerner, R.M., (2002). "Assessment and Recommendations for
Optimal Performance of Waste Containment Systems" U.S. Environmental Protection Agency,
National Risk Management Research Laboratory, Cincinnati, OH (in press).
DRAFT - DO NOT CITE OR QUOTE
A-44
-------
Bourget, S.J., Elrick, D.E., Tanner, C.B. (1958) " Electrical Resistance Units for Moisture
Measurements: Their Moisture Hysterisis, Uniformity, and Sensitivity," Soil Science, Vol 86, pp
298-304.
Bouyoucos, G.J., and Mick, A.H. (1940). "An ElectricalResistance Method for the Continuous
Measurement of Soil Moisture Under Field Conditions" Michigan Agricultural Experiment
Station Bulletin 172, 33 pp.
Cooper, S.P. and Bier, J.D. (1997). "Understanding Landfill Surface Emissions Monitoring",
Proceedings from SWANA's 20th Annual Landfill Gas Symposium, Monterey, California, March ,
pp. 195-213.
Daniel, D.E., Burton, P.M., and Hwang, S.D. (1992). "Evaluation of Four Vadose Zone Probes
Used for Leak Detection and Monitoring," Current Practices in Ground Water and Vadose Zone
Investigations, STP 1118, American Society for Testing and Materials, Philadelphia, Pa., pp.
124-139.
Daniel, D.E., Hamilton, J., and Olson, R.E. (1981). "Suitability of Thermocouple Psychrometers
for Studying Moisture Movement in Unsaturated Soils," Permeability and Groundwater
Containment Transport, STP 764, American Society for Testing and Materials, Philadelphia, Pa.,
pp. 84-101.
Dwyer, S.F. (1997). "Large-Scale Field Study of Landfill Covers at Sandia National
Laboratories," Conference Proceedings: Landfill Capping in the Semi-Arid West: Problems,
Perspectives, and Solutions, T.D. Reynolds and R.C. Morris (eds.), Grand Teton National Park,
Wyoming, Sandi a National Laboratories, Albuquerque, NM, pp. 87-107.
Dwyer, S.F. (1998). "Alternative Landfill Covers Pass the Test", Civil Engineering, ASCE, Vol.
68, No. 9, pp. 50-52.
Dwyer, S.F. (2001). "Finding a Better Cover," Civil Engineering, ASCE, Vol. 71, No. 1, pp 58-
63.
EPA (1991). "Design and Construction of RCRA/CERCLA Final Covers" Seminar Publication,
EPA/625/4-91/025, U.S. Environmental Protection Agency, Office of Research and
Development, Washington, D.C., 145 p.
EPA (1992). "Organic Air Emissions from Waste Management Facilities" Seminar Publication,
EPA/625/R-92/003, U.S. Environmental Protection Agency, Office of Research and
Development, Washington, D.C., 115 p.
Payer, M.J., Rockhold, M.L., and Campbell, M.D. (1992). Use of the Water Balance Method for
Predicting Leachate Generation from Solid Waste Disposal Sites, U.S. Environmental Protection
Agency, EPA/530/SW-168, Washington, D.C., 40 p.
DRAFT - DO NOT CITE OR QUOTE
A-45
-------
Fredlund, D.G. (1992). "Background, Theory, and Research Related to the Use of Thermal
Conductivity Sensors for Matric Suction Measurements," Advance in Measurement of Soil
Physical Properties: Bringing Theory into Practice, SSSA Special Publication No. 30, pp.249-
261.
Gardner, W.R. (1987). "Water Content: An Overview," International Conference on
Measurement of Soil and Plant Water Status, Vol. 1, Utah State University, Loan, Utah, p. 7.
Gee, G.W., Ward, A.L., and Payer, MJ. (1997). "Surface Barrier Research at the Hanford Site,"
Proceedings, International Containment Technology Conference, St. Petersburg, FL., pp. 305-
311.
Lane, D.T., Benson, C.H., Bosscher, P.J., and Pliska, RJ. (1992). "Construction and Hydrologic
Observations of Three Instrumented Final Covers," Proceedings, 15th Annual Madison Waste
Conference, Madison, WI, pp. 231-250.
Lane, D.T., Benson, C.H., and Bosscher, PJ. (1992). "Hydrologic Observations and Modeling
Assessments of Landfill Covers" Environmental Geotechnics Report No. 92-10, Geotechnical
Engineering Program, Department of Civil & Environmental Engineering, University of
Wisconsin, Madison, WI, 406 p.
Lopez, J. and Dwyer, S.F. (1997). "TDR Calibration for the Alternative Landfill Cover
Demonstration (ALCD)." Sandia National Laboratories, Document SAND97-2317-UC-2040.
Kavazanjian, E., Jr. (2000). "Geoenvironmental Performance Monitoring," Performance
Confirmation of Constructed Geotechnical Facilities, Geotechnical Special Publication No. 94,
A.J. Lutenegger and D.J. DeGroot (eds.), ASCE, New York, NY, pp. 21-40.
Khire, M. V. (1995). "Field Hydrology and Water Balance Modeling of Earthen Final Covers
for Waste Containment" Environmental Geotechnics Report No. 95-5, Geotechnical
Engineering Program, Department of Civil & Environmental Engineering, University of
Wisconsin, Madison, WI, 166 p.
Khire, M.V., Benson, C.H., and Bosscher, PJ. (1997). "Water Balance Modeling of Earthen
Final Covers," Journal of Geotechnical and Geoenvironmental Engineering, Vol. 123, No. 8, pp.
744-754.
Khire, M.V., Benson, C.H., and Bosscher, PJ. (1999). "Field Data from a Capillary Barrier and
Model Predictions with UNSAT-H," Journal of Geotechnical and Geoenvironmental
Engineering, Vol. 125, No. 6, pp. 518-527.
Lu, A.-H. and Kunz, C. (1981). "Gas Flow Model to Determine Methane Production at Sanitary
Landfills", Environmental Science and Technology, Vol. 15, pp. 436-440.
Melchior, S. (1997a). "In-situ Studies on the Performance of Landfill Caps (Compacted Soil
Liners, Geomembranes, Geosynthetic Clay Liners, Capillary Barriers)," Proceedings,
International Containment Technology Conference, St. Petersburg, FL, pp. 365-373.
DRAFT - DO NOT CITE OR QUOTE
A-46
-------
Melchior, S. (1997b). "In-situ Studies of the Performance of Landfill Caps (Compacted Soil
Liners, Geomembranes, Geosynthetic Clay Liners, Capillary Barriers)," Land Contamination
and Reclamation, Vol. 5, No. 3, pp. 209-216.
Melchoir, S., Berger, K., Vielhaver, B., and Miehlich, G. (1994). "Multilayered Landfill Covers:
Field Data on the Water Balance and Liner Performance," In-Situ Remediation: Scientific Basis
for Current and Future Technologies, Battelle Press, Columbus, OH, Part 1, pp. 411-425.
Merrill, S.D. and Rawlins, S.L. (1972). "Field Measurement of Soil Water Potential With
Thermocouple Psychrometers," Soil Science, Vol. 112, pp. 102.
Montgomery, RJ. and Parsons, LJ. (1989). "The Omega Hills Final Cover Test Plot Study:
Three-Year Data Summary" presented at Annual Meeting of the National Solid Waste
Management Association, Washington, DC.
Montgomery, RJ. and Parsons, LJ. (1990). "The Omega Hills Final Cover Test Plot Study:
Fourth Year Data Summary" presented at 22nd Mid-Atlantic Industrial Waste Conference,
Drexel University, Philadelphia, PA.
Morisson, R.D. (1983). "Ground Water Monitoring Technology," Timco Manufacturing, Inc.,
Prairie du Sac, WI, pp. 2-31.
Nyhan, J.W., Hakonson, I.E., and Drennon, B J. (1990). "A Water Balance Study of Two
Landfill Cover Designs for Semi-Arid Regions," Journal of Environmental Quality, Vol. 19, No.
2, pp. 281-288.
Nyhan, J.W., Schofield, T.G., and Starmer, R.H. (1997). "Water Balance Study of Four Landfill
Cover Designs Varying in Slope for Semi-Arid Regions," Journal of Environmental Quality,
Vol. 26, No. 5, pp. 1385-1392.
Paige, G.B., Stone, J.J., Lane, LJ., and Hakonson, T.E. (1996). "Calibration and Testing of
Simulation Models for Evaluation of Trench Cap Designs," Journal of Environmental Quality,
Vol. 25, pp. 136-144.
Reece, C.F. (1996). "Evaluation of a Line Heat Dissipation Sensor for Measuring Soil Matric
Potential," Soil Science Society, Am. Journal, Vol. 60, pp. 1022-1028.
Rolston, D.E. (1986). "Gas Flux", Methods of Soil Analysis: Part I, Physical and Mineralogical
Methods, 2nd Edition, Ed. A. Klute, American Society of Agronomy/Soil Science Society of
America, Madison, WI, pp. 1103-1119.
Schofield, T.G., Langhorst, G.J., Trujillo, J., Bostic, K., and Hansen, W. (1994). "Comparison
of Neutron Probe and Time Domain Reflectometry Techniques of Soil Moisture Analysis," Los
Alamos National Laboratory, LA-UR-94-981, Los Alamos, NM.
DRAFT - DO NOT CITE OR QUOTE
A-47
-------
Schultz, R.K., Ridky, R.W., and O'Donnell, E. (1995). "Control of Water Infiltration into Near
Surface LLW Disposal Units, Progress Report of Field Experiments at a Humid Region Site,
Beltsville, Maryland" NUREG/CR-4918, U.S. Nuclear Regulatory Commission, Washington,
DC, Vol. 8, 20 p.
Tanner, C.B., Hanks, RJ. (1952). "Moisture Hysteresis in Gypsum Moisture Blocks," Soil
Science Society Am Proc., Vol.16, pp. 48-51.
Topp, G.C., Davis, J.L., and Annan, A.P. (1980). "Electromagnetic Determination of Soil Water
Content Measurement: Measurement in Coaxial Transmission Lines," Water Resources
Research, Vol. 16, pp. 579.
Webb, S.W., McCord, J.T., and Dwyer, S.F. (1997). "Prediction of Tilted Capillary Barrier
Performance," Proceedings, International Containment Technology Conference, St. Petersburg,
FL, pp. 296-304.
Chapter 9
Collins, P., Ng, A.S., and Ramanujam, R. (1998). "Superfund Success, Superfast," Civil
Engineering, ASCE, Vol. 68, No. 12, pp 58-63.
Dunn, RJ. (1995). "Design and Construction of Foundations Compatible with Solid Wastes,"
Landfill Closures...Environmental Protection and Land Recovery, Geotechnical Special
Publication No. 53, RJ. Dunn and U.P. Singh (eds.), ASCE, New York, NY, pp. 139-159.
EPA (1999a). "Reuse ofCERCLA Landfill and Containment Sites," EPA/540/F-99/015, U.S.
Environmental Protection Agency, Office of Emergency and Remedial Response, Washington,
D.C., 17 p.
EPA (1999b). "Returning Superfund Sites to Productive Use, Bowers Landfill, Pickaway
County, Ohio," EPA/540/F-98/022, U.S. Environmental Protection Agency, Office of
Emergency and Remedial Response, Washington, D.C., 5 p.
EPA(1999c). "Returning Superfund Sites to Productive Use, Chisman Creek, York County,
Virginia," EPA/540/F-98/009, U.S. Environmental Protection Agency, Office of Emergency and
Remedial Response, Washington, D.C., 4 p.
EPA (1999d). "Returning Superfund Sites to Productive Use, Raymark Industries, Inc.,
Stratford, Connecticut" EPA/540/F-98/014, U.S. Environmental Protection Agency, Office of
Emergency and Remedial Response, Washington, D.C., 4 p.
EPA (1999e). "Returning Superfund Sites to Productive Use, Denver Radium, Denver,
Colorado," EPA/540/F-98/010, U.S. Environmental Protection Agency, Office of Emergency
and Remedial Response, Washington, D.C., 5 p.
EPA (200 la). "Reuse Assessments: A Tool to Implement the Superfund Land Use Directive,"
OSWER Directive No. 9355.7-06P, U.S. Environmental Protection Agency, Office of Solid
Waste and Remedial Response, Washington, D.C., 24 p.
DRAFT - DO NOT CITE OR QUOTE
A-48
-------
EPA (200 Ib). "Reusing Superfund Sites: Recreational Use of Land Above Hazardous Waste
Containment Areas" EPA/540/K-01/002, U.S. Environmental Protection Agency, Office of
Emergency and Remedial Response, Washington, D.C., 73 p.
EPA (2002). "Reusing Superfund Sites: Commercial Use Where Waste is Left on Site"
EPA/540/K-01/008, U.S. Environmental Protection Agency, Office of Emergency and Remedial
Response, Washington, D.C., 58 p.
Gross, B.A., Bonaparte, R., and Giroud, J.P. (2002). "Waste Containment Systems: Problems
and Lessons Learned," Appendix F in Assessment and Recommendations for Optimal
Performance of Waste Containment Systems, R. Bonaparte, D.E. Daniel, and R.M. Koerner, U.S.
Environmental Protection Agency, National Risk Management Research Laboratory, Cincinnati,
OH.
Mackey, R.E. (1996). "Three End-Uses for Closed Landfills and Their Impact on the
Geosynthetic Design," Geotextiles and Geomembranes, Vol. 14, Nos. 7/8, pp. 409-424.
DRAFT - DO NOT CITE OR QUOTE
A-49
------- |