United States
Environmental Protection
Agency
Office of
Research and Development
Washington. DC 20460
EPA/625 4-91 '025
May 1991
Seminar Publication
Design and
Construction of
RCRA/CERCLA
Final Covers
-------
Technology Transfer EPA/625/4-91/025
Seminar Publication
Design and Construction
of RCRA/CERCLA
Final Covers
May 1991
Prepared for:
Center for Environmental Research Information
U.S. Environmental Protection Agency
26 West Martin Luther King Drive
Cincinnati, OH 45268
by:
Eastern Research Group, Inc.
6 Whittemore Street
Arlington, MA 02174
Printed on Recycled Paper
-------
NOTICE
The information in this document has been funded wholly or in part by the United States Environmental
Protection Agency under Contract 68-C8-0011 to Eastern Research Group, Inc. It has been subject to
the Agency's peer and administrative review, and it has been approved for publication as an EPA
document. Mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
-------
ACKNOWLEDGMENTS
This seminar publication is based wholly on papers presented at the U.S. Environmental Protec-
tion Agency (EPA) Technology Transfer seminars on Design and Construction of Resource Con-
servation and Recovery Act (RCRA) and Comprehensive Environmental Response,
Compensation, and Liability Act (CERCLA) Final Covers. These seminars were held in July and
August 1990 in Atlanta, Georgia; Philadelphia, Pennsylvania; Boston, Massachusetts; Dallas,
Texas; Kansas City, Missouri; Denver, Colorado; Newark, New Jersey; Chicago, Illinois; Seattle,
Washington; and Oakland, California.
The authors are:
Robert E. Landreth, U.S. Environmental Protection Agency, Risk Reduction Engineering
Laboratory (RREL), Cincinnati, Ohio (Chapters 1 and 5)
Dr. David E. Daniel, Department of Civil Engineering, University of Texas, Austin, Texas
(Chapters 2 and 6)
Dr. Robert M. Koerner, Drexel University, Geosynthetic Research Institute, Philadelphia,
Pennsylvania (Chapters 3, 4, and 7)
Dr. Paul R. Schroeder, Waterways Experiment Station, U.S. Army/Corps of Engineers,
Vicksburg, Mississippi (Chapters 8, 9, and 10)
Dr. Gregory N. Richardson, GN Richardson & Associates, Raleigh, North Carolina (Chapters
11 and 12)
Daniel J. Murray of EPA's Center for Environmental Research Information directed the project,
providing substantive guidance and review. David A. Carson of EPA's Risk Reduction Engineer-
ing Laboratory (RREL), Cincinnati, Ohio, Edwin F. Barth, Jr., of EPA's Center for Environmental
Research Information, Cincinnati, Ohio, and Kenneth R. Skahn of EPA's Office of Solid Waste
and Emergency Response peer reviewed the document. In addition, Frank Walberg, U.S. Army
Corps of Engineers, served as a special reviewer. Susan Richmond, Linda Saunders, Denise
Short, and Heidi Schultz of Eastern Research Group, Inc., provided editorial and production sup-
port.
-------
PREFACE
Cover systems are an essential part of all land disposal facilities. Covers control moisture infiltration
from the surface into closed facilities and limit the formation of leachate and its migration to ground
water. The Resource Conservation and Recovery Act (RCRA) Subparts G, K, and N form the basic
requirements for cover systems being designed and constructed today. In addition, the Comprehen-
sive Environmental Response, Compensation, and Liability Act (CERCLA) refers to RCRA Subtitle
C regulations, and many states have their own more stringent requirements.
This seminar publication provides regulatory and design personnel with an overview of design, con-
struction, and evaluation requirements for cover systems for RCRA/CERCLA waste management
facilities. It offers practical and detailed information on the design and construction of final covers for
both hazardous and nonhazardous waste landfills that comply with these requirements. As such it
should be valuable both to U.S. Environmental Protection Agency (EPA) regional and state person-
nel involved in evaluating and permitting hazardous waste facility closures and to the environmental
design and construction community.
Chapter One presents an overview of cover systems for waste management facilities, including
recommended designs for RCRA Subtitle C and CERCLA facilities. Chapter Two describes soils
used in typical cover systems and discusses critical parameters for soil liners as well as the effects
of environmental impacts such as frost action and settlement. Chapter Three focuses on geosyn-
thetic design and discusses geonet and geocomposite sheet drains, geopipe and geocomposite
edge drains, geotextile filters, geogrid reinforcement, and methane gas vents. Chapter Four covers
durability and aging of geomembranes, discussing in detail the mechanisms of degradation, as well
as synergistic effects, and accelerated testing methods. Chapter Five presents alternative designs
that meet the intent of regulations while adapting to site-specific concerns. Chapter Six discusses
construction quality assurance for soils, including testing of materials, and construction quality as-
surance during all phases of site preparation and soil placement. Chapter Seven covers construc-
tion quality control for geomembranes from manufacture and shipment to placement of the
geomembrane at the site. This chapter also presents destructive and nondestructive tests for sol-
vent and thermal seams in the field. Chapter Eight discusses evaluation of liquid management sys-
tems for landfills using the Hydrologic Evaluation of Landfill Performance (HELP) model. Chapter
Nine examines design parameter effects on cover performance. In Chapter Ten, gas management
systems are discussed with attention to gas generation, migration, and control strategies. Chapter
Eleven presents case studies of five closures, including RCRA industrial and commercial landfills,
one CERCLA lagoon and one CERCLA landfill, and one municipal solid waste commercial landfill.
The final chapter, Chapter Twelve, discusses postclosure monitoring of ground water, leachate, gas
generation, subsidence, surface erosion, and air quality.
This publication is not a design manual nor does it include all of the latest knowledge concerning
RCRA/CERCLA landfill cover systems; additional sources that provide more detailed information are
available. Some of these sources are referred to in the text of the individual chapters. In addition,
state and local authorities should be consulted for regulations and good management practices ap-
plicable to local areas.
IV
-------
TABLE OF CONTENTS
Page
1. OVERVIEW OF COVER SYSTEMS FOR WASTE MANAGEMENT FACILITIES 1
Introduction 1
Recommended Design for Subtitle C Facilities 1
Low Hydraulic Conductivity Layer 2
Compacted Soil Component 2
Geomembrane 2
Drainage Layer 2
Vegetation/Soil Top Layer 3
Vegetation Layer 3
Soil Layer 3
Optional Layers 3
Gas Vent Layer 3
Biotic Layer 4
Subtitle D Covers 4
CERCLA Covers 5
Applicability of RCRA Requirements 5
Relevant and Appropriate RCRA Requirements 6
State Equivalency 6
Closure 6
Applicability of Closure Requirements 6
Relevant and Appropriate Closure Requirements 7
References 7
2. SOILS USED IN COVER SYSTEMS 9
Introduction 9
Typical Cover Systems 9
Flow Rates Through Liners 9
Critical Parameters for Soil Liners 12
Materials : 12
Water Content 13
Compactive Energy 14
Size of Clods 16
Bonding of Lifts 18
Effects of Desiccation 18
Effects of Frost Action 20
Effects of Settlement 20
Interfacial Shear 23
Drainage Layers 24
Summary 25
References ? 25
3. GEOSYNTHETIC DESIGN FOR LANDFILL COVERS 27
General Comments on Design-by-Function 27
-------
Geomembrane Design Concepts 27
Geomembrane Compatibility 27
Vapor Transmission 27
Biaxial Stresses via Subsidence 28
Planar Stresses via Friction 28
Geonet and Geocomposite Sheet Drain Design Concepts 28
Compatibility 28
Crush Strength 28
Flow Capability 29
Geopipe and Geocomposite Edge Drain Design Concepts 30
Compatibility 30
Crush Strength 30
Flow Rate 30
Geotextile Filter Design Considerations 30
Compatibility 30
Permeability 30
Geotextile Soil Retention 32
Geotextile Clogging Evaluation 32
Geogrid, or Geotextile, Cover Soil Reinforcement 32
Geotextile Methane Gas Vent 32
References 33
4. DURABILITY AND AGING OF GEOMEMBRANES 35
Polymers and Foundations 35
Mechanisms of Degradation 35
Ultraviolet Degradation 35
Radiation Degradation 35
Chemical Degradation 35
Swelling Degradation 36
Extraction Degradation 36
Delamination Degradation 36
Oxidation Degradation 36
Biological Degradation 36
Synergistic Effects 37
Elevated Temperature 37
Applied Stresses 37
Long Exposure 37
Accelerated Testing Methods 37
Stress Limit Testing 37
Rate Process Method for Pipe 37
Rate Process Method for Geomembranes 37
Arrhenius Modeling 37
Multi-Parameter Prediction 39
Summary and Conclusions 40
5. ALTERNATIVE COVER DESIGNS 43
Introduction 43
Subtitle C 43
Subtitle D 43
CERCLA 43
Other Cover Designs 44
References 45
6. CONSTRUCTION QUALITY ASSURANCE FOR SOILS 47
Introduction 47
Materials 47
Atterberg Limits 47
VI
-------
Percentage of Fines 47
Percentage of Gravel 47
Maximum Size of Particles or Clods 48
Requirements for Field Personnel 48
Frequency of Testing 48
Control of Subgrade Preparation 48
Soil Placement 48
Soil Compaction 49
Drainage Layers 49
Barrier Materials 49
Protection of a Completed Lift 55
Sampling Pattern 58
Test Pads 58
Outliers 58
Summary 61
Reference 61
7. CONSTRUCTION QUALITY CONTROL FOR GEOMEMBRANES 65
Preliminary Details 65
Manufacture 65
Fabrication of Panels 65
Storage at Factory 65
Shipment 65
Storage at Site 65
Subgrade Preparation 65
Deployment of the Geomembrane 66
Geomembrane Field Seams 66
Solvent Seams 66
Thermal Seams 66
Extrusion Seams 68
Destructive Seam Tests 68
Nondestructive Seam Tests 69
Penetrations, Appurtenances, and Miscellaneous Details 70
Reference 71
8. HYDROLOGIC EVALUATION OF LANDFILL PERFORMANCE (HELP)
MODEL FOR DESIGN AND EVALUATION OF LIQUIDS
MANAGEMENT SYSTEMS 73
Introduction 73
Overview 73
Covers 73
Leachate Collection/Liner Systems 74
HELP Model 75
Background 75
Process Simulation Methods 76
Infiltration 76
Evapotranspiration 77
Subsurface Water Routing 78
Vegetative Growth 79
Accuracy 79
Input Requirements 80
Climatological Data 80
Soil and Design Data 80
Output Description 81
Example Application 81
References 84
VII
-------
9. SENSITIVITY ANALYSIS OF HELP MODEL PARAMETERS 97
Introduction 97
Comparison of Typical Cover Systems 97
Design Parameters 97
Results 98
Effects of Vegetation 98
Effects of Topsoil Thickness 99
Effects of Topsoil Type 102
Use of Lateral Drainage Layer 103
Effects of Climate 103
Vegetative Layer Properties
Effects of SCS Runoff Curve Number 103
Effects of Evaporative Depth 104
Effects of Drainable Porosity 104
Effects of Plant Available Water Capacity 105
Liner/Drain Systems 106
Clay Liner/Drain Systems 107
Geomembrane/Drain Systems 109
Double Liner Systems 111
Summary of Sensitivity Analysis 115
References 116
10. GAS MANAGEMENT SYSTEMS 117
Gas Generation 117
Gas Migration 117
Gas Control Strategies , 118
References 121
11. CASE STUDIES—RCRA/CERCLA CLOSURES 123
Introduction 123
Case 1: RCRA Commercial Landfill 123
Calculation of Localized Subsidence 123
Gas Collection Systems 124
Case 2: RCRA Industrial Landfill 125
CaseS: CERCLA Lagoon Closure 129
Case 4: CERCLA Landfill Closure 132
Case 5: MSW Commercial Landfill 135
Conclusions 138
References 138
Additional References 140
12. POSTCLOSURE MONITORING 141
Introduction , 141
Ground-Water Monitoring 141
Leachate Monitoring 141
Gas Generation 143
Subsidence Monitoring 144
Surface Erosion 145
Air Quality Monitoring 145
References 145
Appendix A
Stability and Tension Considerations Regarding Cover Soils on Geomembrane-Lined Slopes A-1
Appendix B
Long-term Durability and Aging of Geomembranes B-1
VIII
-------
L/STOFF/G(//?ES
Figure Page
1-1 EPA-recommended landfill cover design 2
1-2 EPA-recommended landfill cover with options 4
2-1 Soil liner, geomembrane liner, and composite liner 10
2-2 Soil liner and composite liner 12
2-3 Effect of bentonite upon the hydraulic conductivity of a bentonite-amended soil 14
2-4 Hydraulic conductivity and dry unit weight versus molding water content 15
2-5 Highly plastic soil compacted with standard Proctor procedures at a water content of 12% 16
2-6 Highly plastic soil compacted with standard Proctor procedures at a water content of 16% 16
2-7 Highly plastic soil compacted with standard Proctor procedures at a water content of 20% 16
2-8 Rototiller used to mix soil 17
2-9 Blades and teeth on rototiller 17
2-10 Influence of compactive effort upon hydraulic conductivity and dry unit weight 19
2-11 Road recycler used to pulverize clods of soil 20
2-12 Passage of road recycler over loose lift of mudstoneto reduce size of chunks of mudstone 21
2-13 Effect of imperfect bonding of lifts on hydraulic performance of soil liner 21
2-14 Example of heavy footed roller with long feet 22
2-15 Effect of desiccation upon the hydraulic conductivity of compacted clay 23
2-16 Relationship between distortion and tensile strain 24
2-17 Relationship between shearing characteristics of compacted soils and conditions of compaction 25
3-1 Required strength 28
3-2 Response of common geomembranes to the three-dimensional geomembrane tension test 29
3-3 Required geomembrane tension 29
3-4 Common crush strength behavior for geonets and geocomposites 30
3-5 ASTM D-4716 flow rate test 31
3-6 Crush strength of geopipe and geocomposite edge drain cores 32
3-7 Required strength of geogrid for cover soil reinforcement 33
3-8 Allowable gas flow as adapted from ASTM D-4716 34
4-1 Wavelength spectrum of visible and ultraviolet radiation 36
4-2 Stress limit testing for plastic pipe 38
4-3 Rate process method for testing pipe 38
4-4 Rate process method for testing geomembranes 39
4-5 Testing device for Arrhenius modeling 39
4-6 Reaction rate for impact testing of polyethylene shielding 40
4-7 Experimental and field-measured response curves for multi-parameter lifetime prediction 41
5-1 Resistive layer barrier 44
5-2 Conductive layer barrier 45
5-3 Side view of bioengineered lysimeter. Surface runoff is collected from both engineered surface and
soil surface. Soil moisture content is measured with neutron probe. Water table is measured
in well 45
6-1 Traditional method for specification of acceptable water contents and dry unit weights 50
6-2 Data from Mitchell et al. for silty clay compacted with impact compaction 51
6-3 Compaction data for silty clay (6); solid symbols represent specimens with hydraulic conductivity less
than or equal to 1 x 10"7 cm/s and open symbols represent specimens with hydraulic conductivity
>1 x 10'7cm/s 52
6-4 Contours of constant hydraulic conductivity for silty clay compacted with kneading compaction 52
6-5 Recommended procedure 53
6-6 Use of hydraulic conductivity and shear strength data to define a single, overall acceptable zone 55
6-7 Possible approaches for specifying lower limit of Acceptable Zone: (A) minimum degree of
saturation, S; and (B) line of optimums 56
6-8 Compaction curves for Type A soil from East Borrow area at Oak Ridge Y-12 operations project 57
6-9 Hydraulic conductivity versus molding water content for Type A soil from East Borrow area at Oak
Ridge Y-12 operations project 57
6-10 Acceptable zone for Type A soil from East Borrow area at Oak Ridge Y-12 operations project 58
IX
-------
6-11 Pushing of thin-walled sampling tube with abackhoe 59
6-12 Tilting of sampling tube during push 59
6-13 Placement of hydraulic jack on top of sampling tube 60
6-14 Use of backhoe as a reaction for hydraulic jack 60
6-15 Checklist of critical variables for CQA of low hydraulic conductivity compacted soil used in a
cover system 62
6-16 Checklist of critical variables for CQA of drainage materials used in a cover system 63
7-1 Shear and peel test for geomembrane seams 68
8-1 Cover and liner edge configuration with example toe drain 73
8-2 Schematic of a single clay liner system for a landfill 74
8-3 Schematic of a double liner and leak detection system for a landfill 75
8-4 Simulation processes in the HELP model 76
8-5 Typical hazardous waste landfill profile 81
8-6 Completed data form for landfill materials and design 82
8-7 Completed data form for climatological data 84
8-8 Example output 85
9-1 Cover designs for sensitivity analysis 99
9-2 Bar graph for three-layer cover design showing effect of surface vegetation, topsoil type,
and location 100
9-3 Bar graph for two-layer cover design showing effect of topsoil depth, surface vegetation,
and location 100
9-4 Effect of saturated hydraulic conductivity on lateral drainage and percolation 109
9-5 Effect of ratio of drainage layer saturated hydraulic conductivity to soil liner saturated hydraulic
conductivity on ratio of lateral drainage to percolation for a steady-state (SS) inflow of
20 cm/yr(8 in./yr) 110
9-6 Effect of ratio of drainage layer saturated hydraulic conductivity to soil liner saturated hydraulic
conductivity on ratio of lateral drainage to percolation for an unsteady inflow of
127 cm/yr (50 in./yr) 110
9-7 Effect of ratio of drainage length to drainage layer slope on the average saturated depth in
drainage layer (KD=10~2 cm/s) above a soil liner (KP=10~7cm/s) under a steady-state inflow
rate of 20 cm/yr(8 in./yr) 111
9-8 Synthetic liner leakage fraction as a function of density of holes, size of holes, head on the liner
and saturated hydraulic conductivity of the liner 112
9-9 Effect of leakage fraction on system performance 112
9-10 Liner designs 113
9-11 Percent of inflow to primary leachate collection layer discharging from leakage detection layer and
bottom liner for double liner systems C and E 114
9-12 Percent of inflow to primary leachate collection layer discharging from leakage detection layer and
bottom liner for double liner systems Dand F 114
10-1 Cover with gas vent outlet and vent layer 118
10-2 Gravel vent and gravel-filled trench used to control lateral gas movement in a sanitary landfill 119
10-3 Typical trench barrier system 119
10-4 Gas control barriers 120
10-5 Gas extraction well for landfill gas control 121
10-6 Gas extraction well design 122
11-1 Case 1 - Cap profile and geometry 124
11-2 Case 1 - General subsidence model 125
11-3 Case 1 - Cumulative subsidence 126
11-4 Case 1 - Geomembrane strains intrench subsidence 126
11-5 Case 1 - Uniaxial and biaxial geomembrane response 127
11-6 Case 1 - Subsidence strain in soil barrier 127
11-7 Case 1 - Ultimate tensile strain in clays 128
11-8 Case 1 - Gas collector system 129
11-9 Case 2 - Cap profile and geometry 130
11-10 Case 2- Direct shear data: texture HOPE 130
11-11 Case 2 - Slope factors for soil loss evaluation 131
11-12 Case 2 - Sideslope armoring scheme 131
-------
11-13 Case 3 - Cap profile and geometry 132
11-14a Cases- Placement of geogrid over geomembrane 133
11-14b CaseS - Placement of drainage layer over geogrid 133
11-15a Case 3 - Outlet detail for sideslope toe surface water drainage layer 134
11-15b Case 3 - Erosion at drainage layer outlet 134
11-16 Case 4 - Cap profile and geometry 135
I1-17a Case 4 - Placement of geotextile on asphalt emulsion 136
11-17b Case 4 - Placement of chip seal on geotextile 136
11-18 Case 5 - Cap profile and geometry 137
11-19 CaseS- Profile showing MSW subcells 138
11-20 Case 5 - Gas collector well array 139
11-21 Case 5 - Perimeter gas monitoring well 139
12-1 Monitoring well configuration 142
12-2 Monitoring interbedded aquifer 142
12-3 Impact of biological growth on filters 143
12-4 Gas generation versus time 143
XI
-------
LIST OF TABLES
Table Page
2-1 Calculated Flow Rates through Soil Liners with 30 cm of Water Ponded on the Liner 10
2-2 Calculated Flow Rates through a Geomembrane with a Head of 30 cm of Water above
the Geomembrane 10
2-3 Calculated Flow Rates for Composite Liners with a Head of Water of 30 cm 11
2-4 Calculated Flow Rates for Soil Liners, Geomembrane Liners, and Composite Liners 13
2-5 Effect of Size of Clods during Processing of Soil upon Hydraulic Conductivity of Soil
after Compaction 18
3-1 Customary Primary Functions of Geosynthetics Used in Waste Containment Systems 27
4-1 Typical Formulations of Geomembranes 35
6-1 Recommended Materials Tests for Barrier Layers 48
6-2 Recommended Tests and Observations on Subgrade Preparation 49
6-3 Recommended Tests and Observations on Compacted Soil for Barrier Layers 54
7-1 Overview of Geomembrane Field Seams 67
7-2 Overview of Nondestructive Seam Tests 69
9-1 Parameters Selected for Sensitivity Analysis 97
9-2 Climatological Regimes 98
9-3 Effects of Climate and Vegetation 101
9-4 Effects of Climate and Topsoil Thickness 101
9-5 Effects of Climate and Topsoil Types 102
9-6 Effects of Evaporative Depth and Runoff Curve Number 105
11-1 Soil Texture Constant for Soil Loss Evaluation 131
12-1 Threshold Limits of Air Contamination 144
XII
-------
CHAPTER 1
OVERVIEW OF COVER SYSTEMS FOR WASTE MANAGEMENT FACILITIES
INTRODUCTION
Proper closure is essential to complete a filled hazardous
waste landfill. Research has established minimum re-
quirements needed to meet the stringent, necessary,
closure regulations in the United States. In designing the
landfill cover, the objective is to limit the infiltration of
water to the waste so as to minimize creation of leachate
that could possibly escape to ground-water sources.
Minimizing leachates in a closed waste management unit
requires that liquids be kept out and that the leachate
that does exist be detected, collected, and removed.
Where the waste is above the ground-water zone, a
properly designed and maintained cover can prevent (for
practical purposes) water from entering the landfill and,
thus, minimize the formation of leachate.
The cover system must be devised at the time the site is
selected and the plan and design of the landfill contain-
ment structure is chosen. The location, the availability of
soil with a low permeability or hydraulic conductivity, the
stockpiling of good topsoil, the availability and use of
geosynthetics to improve performance of the cover sys-
tem, the height restrictions to provide stable slopes, and
the use of the site after the postclosure care period are
typical considerations. The goals of the cover system are
to minimize further maintenance and to protect human
health and the environment.
Subparts G, K, and N of the Resource Conservation and
Recovery Act (RCRA) Subtitle C regulations form the
basic requirements for cover systems being designed
and constructed today. Comprehensive Environmental
Response, Compensation, and Liability Act (CERCLA)
regulations refer to the RCRA Subtitle C regulations but
other criteria, primarily approved state requirements, also
have to be evaluated for applicability. The proposed
RCRA Subtitle D regulations base cover requirements
primarily on the hydraulic conductivity of the bottom liner.
RECOMMENDED DESIGN FOR SUBTITLE C
FACILITIES
After the hazardous waste management unit is closed,
the U.S. Environmental Protection Agency (EPA) recom-
mends (1) that the final cover (Figure 1-1) consist of,
from bottom to top:
1. A Low Hydraulic Conductivity Geomembrane/Soil
Layer. A 60-cm (24-in.) layer of compacted natural or
amended soil with a hydraulic conductivity of 1 x 10"7
cm/sec in intimate contact with a minimum 0.5-mm
(20-mil) geomembrane liner.
2. A Drainage Layer. A minimum 30-cm (12-in.) soil
layer having a minimum hydraulic conductivity of 1 x
10~2 cm/sec, or a layer of geosynthetic material
having the same characteristics.
3. A Top, Vegetation/Soil Layer. A top layer with
vegetation (or an armored top surface) and a mini-
mum of 60 cm (24 in.) of soil graded at a slope bet-
ween 3 and 5 percent.
Because the design of the final cover must consider the
site, the weather, the character of the waste, and other
site-specific conditions, these minimum recommenda-
tions may be altered providing the alternative design is
equivalent to the EPA-recommended design or will meet
the intent of the regulations. EPA encourages design in-
novation and will accept an alternative design provided
the owner or operator demonstrates the new design's
equivalency. For example, in extremely arid regions, a
gravel top surface might compensate for reduced vegeta-
tion, or the middle drainage layer might be expendable.
Where burrowing animals might damage the
geomembrane/low hydraulic conductivity soil layer, a
biotic barrier layer of large-sized cobbles may be needed
above it. Where the type of waste may create gases, soil
or geosynthetic vent structures would need to be in-
cluded.
Settlement and subsidence should be evaluated for all
covers and accounted for in the final cover plans. The
current operating procedures for RCRA Subtitle C
facilities (e.g., banning of liquids and partially filled drums
of liquids) usually do not present major settlement or sub-
sidence issues. For RCRA Subtitle D facilities, however,
the normal decomposition of the waste will invariably
result in settlement and subsidence. Settlement and sub-
sidence can be significant, and special care may be re-
quired in designing the final cover system. The cover
design process should consider the stability of all the
waste layers and their intermediate soil covers, the soil
and foundation materials beneath the landfill site, all the
-------
vegetation/soil
top layer
drainage layer
low hydraulic conductivity
geomembrane/soil layer
waste
\1// \\// \1// \1// \\//
O
0
O
Figure 1-1. EPA-recommended landfill cover design (1).
liner and leachate collection systems, and all the final
cover components. When a significant amount of settle-
ment and subsidence is expected within 2 to 5 years of
closure, an interim cover that protects human health and
the environment might be proposed. Then when settle-
ment/subsidence is essentially complete, the interim
cover could be replaced or incorporated into a final cover.
Low Hydraulic Conductivity Layer
The function of the composite low hydraulic conductivity
layer, composed of soil and a geomembrane, is to
prevent moisture movement downward from the overlying
drainage layer.
Compacted Soil Component
EPA recommends a test pad be constructed before the
low hydraulic conductivity soil layer is put in place to
demonstrate that the compacted soil component can
achieve a maximum hydraulic conductivity of 1 x 10~7
cm/sec. To ensure that the design specifications are at-
tainable, a test pad uses the same soil, equipment, and
procedures to be used in constructing the low hydraulic
conductivity layer. For Subtitle D facilities, the test fill
should be constructed on part of the solid waste material
to determine the impact of compacting soil on top of less
resistive municipal solid waste.
The low hydraulic conductivity soil component placed
over the waste should be at least 60-cm (24-in.) deep;
free of detrimental rock, clods, and other soil debris; have
an upper surface with a 3 percent maximum slope; and
be below the maximum frost line. The surface should be
smooth so that no small-scale stress points are created
for the geomembrane.
In designing the low hydraulic conductivity layer, the
causes of failure—subsidence, desiccation cracking, and
freeze/thaw cycling—must be considered. Most of the
settling will have taken place by the time the cover is put
into place, but there is still a potential for further sub-
sidence. Although estimating this potential is difficult, in-
O
0
0
03
O
60cm
' -^— filter layer
30 cm
*»- 0.5-mm (20-mil)
60 cm geomembrane
formation about voids and compressible materials in the
underlying waste will aid in calculating subsidence.
A soil with low cracking potential should be selected for
the soil component of the low hydraulic conductivity layer.
The potential for desiccation cracking of compacted clay
depends on the physical properties of the compacted
clay, its moisture content, the local climate, and the mois-
ture content of the underlying waste.
Because freeze/thaw conditions can cause soil cracking,
lessen soil density, and lessen soil strength, this entire
low hydraulic conductivity/geomembrane layer should be
below the depth of the maximum frost penetration. In
northern areas, then, the maximum depth of the top
vegetation/soil layer would be greater than the recom-
mended minimum of 60 cm (24 in.).
Penetrating this low hydraulic conductivity/geomembrane
soil layer with gas vents or drainage pipes should be kept
to a minimum. Where a vent is necessary, there should
be a secure, liquid-tight seal between the vent and the
geomembrane. If settlement or subsidence is a major
concern, this seal must be designed for flexibility to allow
for vertical movement.
Geomembrane
The geomembrane placed on the smooth, even, low
hydraulic conductivity layer should be at least 0.5-mm
(20-mils) thick. The minimum slope surface should be 3
percent after any settlement of the soil layer or sub-base
material. Stress situations such as bridging over sub-
sidence and friction between the geomembrane and
other cover components (i.e., compacted soil, geosyn-
thetic drainage material, etc.), especially on side slopes,
will require special laboratory tests to ensure the design
has incorporated site-specific materials.
Drainage Layer
The drainage layer should be designed to minimize the
time the infiltrated water is in contact with the bottom, low
hydraulic conductivity layer and, hence, to lessen the
-------
potential for the water to reach the waste (see Figure 1-
1). Water that filters through the top layer is intercepted
and rapidly moved to an exit drain, such as by gravity
flow to a toe drain.
If the granular material in the drainage layer is sand, the
minimum requirements are that it should be at least 30-
cm (12-in.) deep with a hydraulic conductivity of 1 x 10~2
cm/sec or greater. Drainage pipes should not be placed
in any manner that would damage the geomembranes.
If geosynthetic materials are used in the drainage layer,
the same physical and hydraulic requirements should be
met, e.g., equivalency in hydraulic transmissivity, lon-
gevity, compatibility with geomembrane, compressibility,
conformance to surrounding materials, and resistance to
clogging. Geosynthetic materials are gaining increased
use and understanding of their performance. Manufac-
turers are also continuing to improve the basic resin
properties to improve their long-term durability. The net
result is that organizations such as the American Society
of Testing Materials (ASTM) and the Geosynthetic Re-
search Institute (GRI), Drexel University, Philadelphia,
Pennsylvania, are continually developing new evaluation
procedures to better correlate with design and field ex-
periences.
Between the bottom of the top-layer soil and the
drainage-layer sand, a granular or geosynthetic filter
layer should be included to prevent the drainage layer
from clogging by top-layer fines. The criteria established
for the grain size of granular filter sand are designed to
minimize the migration of fines from the overlying top
layer into the drainage layer. (For information on filter
criteria, refer to the EPA Technical Guidance Document
[1].) ASTM test procedures have also been established
to evaluate paniculate clogging potential of geosyn-
thetics.
Vegetation/Soil Top Layer
Vegetation Layer
The upper layer of the two-component top layer (Figure
1-1) should be vegetation (or another surface treatment)
that will allow runoff from major storms while inhibiting
erosion. Vegetation over soil (part of which is topsoil) is
the preferred system, although, in some areas, vegeta-
tion may be unsuitable.
The temperature- and drought-resistant vegetation
should be indigenous; have a root system that does not
extend into the drainage layer; need no maintenance;
survive in low-nutrient soil; and have sufficient density to
control the rate of erosion to the recommended level of
less than 5.5 MT/ha/yr (2 ton/acre/yr).
The surface slope should be the same as that of the un-
derlying soils; at least 3 percent but no greater than 5
percent. To support the vegetation, this top layer should
be at least 60-cm (24-in.) deep and include at least 15-
cm (6-in.) of topsoil. To help the plant roots develop, this
layer should not be compacted. In some northern
climates, this top layer may need to be more than the
minimum 60 cm (24 in.) to ensure that the bottom low
hydraulic conductivity layer remains below the frost zone.
Where vegetation cannot be maintained, particularly in
arid areas, other materials should be selected to prevent
erosion and to allow for surface drainage. Asphalt and
concrete are apt to deteriorate because of thermal-
caused cracking or deform because of subsidence.
Therefore, a surface layer 13 to 25-cm (5 to 10-in.) deep
of 5 to 10-cm (2 to 4-in.) stones or cobbles would be
more effective. Although cobbles are a one-way valve
and allow rain to infiltrate, this phenomenon would be of
less concern in arid areas. In their favor, cobbles resist
wind erosion well.
Soil Layer
The soil in this 60-cm (24-in.) top layer should be capable
of sustaining nonwoody plants, have an adequate water-
holding capacity, and be sufficiently deep to allow for ex-
pected, long-term erosion losses. A medium-textured soil
such as a loam would fit these requirements. If the landfill
site has sufficient topsoil, it should be stockpiled during
excavation for later use.
The final slopes of the cover should be uniform and at
least 3 percent, and should not allow erosion rills and gul-
lies to form. Slopes greater than 5 percent will promote
erosion unless controls are built in to limit erosion to less
than 5.5 MT/ha/yr (2 ton/acre/yr). The U.S. Department of
Agriculture's (USDA's) Universal Soil Loss Equation is
recommended as the tool to evaluate erosion potential.
Optional Layers
Although other layers may be needed on a site-specific
basis, the common optional layers are those for gas
vents and for a biotic barrier layer (Figure 1 -2).
Gas Vent Layer
The gas vent layer should be at least 30-cm (12-in.) thick
and be above the waste and below the low hydraulic con-
ductivity layer. Coarse-grained porous material, similar to
that used in the drainage layer or equivalent-performing
synthetic material, can be used.
The perforated, horizontal venting pipes should channel
gases to a minimum number of vertical risers located at a
high point (in the cross section) to promote gas ventila-
tion. To prevent clogging, a granular or geotextile filter
may be needed between the venting and the low
hydraulic conductivity soil geomembrane layers.
As an alternative, vertical, standpipe gas collectors can
be built up as the landfill is filled with waste. These
standpipes, which may be constructed of concrete, can
be 30 cm (12 in.) or more in diameter and may also be
used to provide access to measure leachate levels in the
landfill.
-------
cobbles/soil
top layer
biotic barrier
(cobbles)
drainage layer
low hydraulic conductivity
geomembrane/soil layer
gas vent layer
waste
o o O
o C, o
o 0 0 o
<0
r>°
0>
°
60cm
30cm
30cm
geosynthetic filter
geosynthetic filter
0.5-mm (20 mil)
geomembrane
geosynthetic filter
Figure 1-2. EPA-recommended landfill cover with options (1).
Biotic Layer
Plant roots or burrowing animals (collectively called
biointruders) may disrupt the drainage and the low
hydraulic conductivity layers to interfere with the drainage
capability of the layers. A 90-cm (3-ft.) biotic barrier of
cobbles directly beneath the top vegetation layer may
stop the penetration of some deep-rooted plants and the
invasion of burrowing animals. Most research on biotic
barriers has been done in, and is applicable to, arid
areas. Geosynthetic products that incorporate a time-
released herbicide into the matrix or on the surface of the
polymer may also be used to retard plant roots. The lon-
gevity of these products requires evaluation if the cover
system is to serve for longer than 30 to 50 years.
SUBTITLE D COVERS
The cover system in nonhazardous waste landfills (Sub-
title D) will be a function of the bottom liner system and
the liquids management strategy for the specific site. If
the bottom liner system contains a geomembrane, then
the cover system should contain a geomembrane to
prevent the "bathtub" effect. When the bottom liner is less
permeable than the cover system, e.g., geomembrane on
the bottom and natural soil on the top, the facility will "fill
up" with infiltration water (through the cover) unless an
active leachate removal system is in place. Likewise, if
the bottom liner system is a natural soil liner, then the
cover system barrier should be hydraulically equivalent to
or less than the bottom liner system. A geomembrane
used in the cover will prevent the infiltration of moisture to
the waste below and may contribute to the collection of
waste decomposition gases, therefore necessitating a
gas-vent layer.
There are at least two options to consider under a liquids
management strategy, mummification and recirculation.
In the mummification approach the cover system is
designed, constructed, and maintained to prevent mois-
ture infiltration to the waste below. The waste will even-
tually approach and remain in a state of "mummification"
until the cover system is breached and moisture enters
the landfill. A continual maintenance program is neces-
sary to maintain the cover system in a state of good
repair so that the waste does not decompose to generate
leachate and gas.
The recirculation concept results in the rapid physical,
chemical, and biological stabilization of the waste. To ac-
complish this, a moisture balance is maintained within
the landfill that will accelerate these stabilization proces-
ses. This approach requires geomembranes in both the
bottom and top control systems to prevent leachate from
getting out and excess moisture from getting in. In addi-
tion, the system needs a leachate collection and removal
system on the bottom and a leachate injection system on
the top, maintenance of this system for a number of
years (depending on the size of the facility), and a gas
collection system to remove the waste decomposition
gases. In a modern landfill facility, all of these elements,
except the leachate injection system, would probably be
available. The benefit of this approach is that, after
stabilization, the facility should not require further main-
tenance. A more important advantage is that the decom-
posed and stabilized waste may be removed and used
like compost, the plastics and metals could be recycled,
and the site used again. If properly planned and operated
in this manner, several cells could serve all of a
community's waste management needs.
A natural soil material may be used in a cover system
when the bottom liner system is also natural soils and the
regulatory requirements will permit. A matrix of soil char-
acteristics (using either USDA or USCS) and health, aes-
thetics, and site usage characteristics can be developed
to provide information on which soil or combination of
soils will be the most beneficial.
Health considerations demand the evaluation of each soil
type to minimize vector breeding areas and attractive-
ness to animals. The soil should minimize moisture in-
-------
filtration (best accomplished by fine grain soils) while al-
lowing gas movement (coarse grain soils are best). This
desired combination of seemingly opposite soil properties
suggests a layered system. The soil should also minimize
fire potential.
Aesthetic considerations include minimizing blowing of
paper and other waste, controlling odors, and providing a
sightly appearance. All landfill operators strive to be good
neighbors and these considerations are very important
for community relations.
The landfill site may be used for a variety of activities
after closure. For this reason, cover soils should minimize
settlement and subsidence, maximize compaction, assist
vehicle support and movement, allow for equipment
workability under all weather conditions, and allow heal-
thy vegetation to grow. The future use of the site should
be considered at the initial landfill design stages so that
appropriate end-use design features can be incorporated
into the cover during the active life of the facility.
CERCLA COVERS
The Superfund Amendments and Reauthorization Act of
1986 (SARA) adopts and expands a provision in the
1985 National Contingency Plan (NCR) that remedial ac-
tions must at least attain applicable or relevant and ap-
propriate requirements (ARARs). Section 121 (d) of
CERCLA, as amended by SARA, requires attainment of
federal ARARs and of state ARARs in state environmen-
tal or facility siting laws when the state requirements are
promulgated, more stringent than federal laws, and iden-
tified by the state in a timely manner.
CERCLA facilities require information on whether or not
the site is under the jurisdiction of RCRA regulations. The
cover system design can then be developed based on
appropriate regulations.
RCRA Subtitle C requirements for treatment, storage,
and disposal facilities (TSDFs) will frequently be ARARs
for CERCLA actions, because RCRA regulates the same
or similar wastes as those found at many CERCLA sites,
covers many of the same activities, and addresses
releases and threatened releases similar to those found
at CERCLA sites. When RCRA requirements are ARARs,
only the substantive requirements of RCRA must be met
if a CERCLA action is to be conducted on site. Substan-
tive requirements are those requirements that pertain
directly to actions or conditions in the environment. Ex-
amples include performance standards for incinerators
(40 CFR 264.343), treatment standards for land disposal
of restricted waste (40 CFR 268), and concentration
limits, such as maximum contaminant levels (MCLs). On-
site actions do not require RCRA permits or compliance
with administrative requirements. Administrative require-
ments are those mechanisms that facilitate the im-
plementation of the substantive requirements of a statute
or regulation. Examples include the requirements for
preparing a contingency plan, submitting a petition to
delist a listed hazardous waste, recordkeeping, and con-
sultations. CERCLA actions to be conducted off site must
comply with both substantive and administrative RCRA
requirements.
APPLICABILITY OF RCRA REQUIREMENTS
RCRA Subtitle C requirements for the treatment, storage,
and disposal of hazardous waste are applicable for a Su-
perfund remedial action if the following conditions are
met (2):
1. The waste is a RCRA hazardous waste, and either:
2. The waste was initially treated, stored, or disposed of
after the effective date of the particular RCRA re-
quirement
or
The activity at the CERCLA site constitutes treat-
ment, storage, or disposal, as defined by RCRA.
For RCRA requirements to be applicable, a Superfund
waste must be determined to be a listed or characteristic
hazardous waste under RCRA. A waste that is hazard-
ous because it once exhibited a characteristic (or a
media containing a waste that once exhibited a charac-
teristic) will not be subject to Subtitle C regulation if it no
longer exhibits that characteristic. A listed waste may be
delisted if it can be shown not to be hazardous based on
the standards in 40 CFR 264.22. If such a waste will be
shipped off site, it must be delisted through a rulemaking
process. To delist a RCRA hazardous waste that will
remain on site at a Superfund site, however, only the
substantive requirements for delisting must be met.
Any environmental media (i.e., soil or ground water) con-
taminated with a listed waste is not a hazardous waste,
but must be managed as such until it no longer contains
the listed waste—generally when constituents from the
listed waste are at health-based levels. Delisting is not
required.
To determine whether a waste is a listed waste under
RCRA, it is often necessary to know the source of that
waste. For any Superfund site, if determination cannot be
made that the contamination is from a RCRA hazardous
waste, RCRA requirements will not be applicable. This
determination can be based on testing or on best profes-
sional judgment (based on knowledge of the waste and
its constituents).
A RCRA requirement will be applicable if the hazardous
waste was treated, stored, or disposed of after the effec-
tive date of the particular requirement. The RCRA Sub-
title C regulations that established the hazardous waste
management system first became effective on November
19, 1980. Thus, RCRA regulations will not be applicable
to wastes disposed of before that date, unless the
CERCLA action itself constitutes treatment, storage, or
disposal (see below). Additional standards have been is-
-------
sued since 1980; therefore, applicable requirements may
vary somewhat, depending on the specific date on which
the waste was disposed.
RCRA requirements for hazardous wastes will also be
applicable if the response activity at the Superfund site
constitutes treatment, storage, or disposal, as defined
under RCRA. Because remedial actions frequently in-
volve grading, excavating, dredging, or other measures
that disturb contaminated material, activities at Superfund
sites may constitute disposal, or placement, of hazardous
waste. Disposal of hazardous waste, in particular, trig-
gers a number of significant requirements, including
closure requirements and land disposal restrictions,
which require treatment of wastes prior to land disposal.
(See Guides on Superfund Compliance with Land Dis-
posal Restrictions, OSWER Directives 9347.3-01 FS
through 9237.3-06FS, for a detailed description of these
requirements.)
EPA has determined that disposal occurs when wastes
are placed in a land-based unit. However, movement
within a unit does not constitute disposal or placement,
and at CERCLA sites, an area of contamination (AOC)
can be considered comparable to a unit. Therefore,
movement within an AOC does not constitute placement.
Relevant and Appropriate RCRA Requirements
RCRA requirements that are not applicable may, none-
theless, be relevant and appropriate, based on site-
specific circumstances. For example, if the source or
prior use of a CERCLA waste is not identifiable, but the
waste is similar in composition to a known, listed RCRA
waste, the RCRA requirements may be potentially
relevant and appropriate, depending on other circumstan-
ces at the site. The similarity of the waste at the CERCLA
site to RCRA waste is not the only, nor necessarily the
most important, consideration in the determination. An in-
depth, constituent-by-constituent analysis is generally
neither necessary nor useful, since most RCRA require-
ments are the same for a given activity or unit, regardless
of the specific composition of the hazardous waste.
The determination of relevance and appropriateness of
RCRA requirements is based instead on the circumstan-
ces of the release, including the hazardous properties of
the waste, its composition and matrix, the characteristics
of the site, the nature of the release or threatened
release from the site, and the nature and purpose of the
requirement itself. Some requirements may be relevant
and appropriate for certain areas of the site, but not for
other areas. In addition, some RCRA requirements may
be relevant and appropriate at a site, while others are
not, even for the same waste. For example, at one site
minimum technology requirements may be considered
relevant and appropriate for an area receiving waste be-
cause of the high potential for migration of contaminants
in hazardous levels to ground water, but not for another
area that contains relatively immobile waste. Land dis-
posal restrictions at the same site may not be relevant
and appropriate for either area because the required
treatment technology is not appropriate, given the matrix
of the waste. Only those requirements that are deter-
mined to be both relevant and appropriate must be at-
tained.
State Equivalency
A state may be authorized to administer the RCRA haz-
ardous waste program in lieu of the federal program
provided the state has equivalent authority. Authorization
is granted separately for the basic RCRA Subtitle C
program, which includes permitting and closure of
TSDFs; for regulations promulgated pursuant to the Haz-
ardous and Solid Waste Amendments (HSWA), such as
land disposal restrictions; and for other programs, such
as delisting of hazardous wastes. If a site is located in a
state with an authorized RCRA program, the state's
promulgated RCRA requirements will replace the
equivalent federal requirements as potential ARARs.
An authorized state program may also be more stringent
than the federal program. For example, a state may have
more stringent test methods for characteristic wastes, or
may list more wastes as hazardous than the federal
program does. Therefore, it is important to determine
whether laws in an authorized state go beyond the
federal regulations.
Closure
For each type of unit regulated under RCRA, Subtitle C
regulations contain standards that must be met when a
unit is closed. For treatment and storage units, the
closure standards require that all hazardous waste and
hazardous waste residues be removed. In addition to the
option of closure by removal, called clean closure, units
such as landfills, surface impoundments, and waste piles
may be closed as disposal or landfill units with waste in
place, referred to as landfill closure. Frequently, the
closure requirements for such land-based units will be
either applicable or relevant and appropriate at Super-
fund sites.
Applicability of Closure Requirements
The basic prerequisites for applicability of closure re-
quirements are (1) the waste must be hazardous waste;
and (2) the unit (or AOC) must have received waste after
the RCRA requirements became effective, either be-
cause of the original date of disposal or because the
CERCLA action constitutes disposal. When RCRA
closure requirements are applicable, the regulations
allow only two types of closure:
• Clean Closure. All waste residues and contaminated
containment system components (e.g., liners), con-
taminated subsoils, and structures and equipment con-
taminated with waste leachate must be removed and
managed as hazardous waste or decontaminated
before the site management is completed [see 40 CFR
264.111,264.228(3)].
-------
• Landfill Closure. The unit must be capped with a final
cover designed and constructed to:
• Provide long-term minimization of migration of li-
quids.
• Function with minimum maintenance.
• Promote drainage and minimize erosion.
• Accommodate settling and subsidence.
• Have a hydraulic conductivity less than or equal to
any bottom liner system or natural subsoils present.
Clean closure standards assume the site will have un-
restricted use and require no maintenance after the
closure has been completed. These standards are often
referred to as the "eatable solid, drinkable leachate"
standards. In contrast, disposal or landfill closure stand-
ards require postclosure care and maintenance of the
unit for at least 30 years after closure. Postclosure care
includes maintenance of the final cover, operation of a
leachate and removal system, and maintenance of a
ground-water monitoring system [see 40 CFR 264.117,
264.228(b)].
EPA has prepared several guidance documents on
closure and final covers (1, 3). These guidance docu-
ments are not ARARs, but are to be considered for
CERCLA actions and may assist in complying with these
regulations. The performance standards in the regulation
may be attained in ways other than those described in
guidance, depending on the specific circumstances of the
site.
Relevant and Appropriate Closure Requirements
If they are not applicable, RCRA closure requirements
may be determined to be relevant and appropriate. There
is more flexibility in designing closure for relevant and ap-
propriate requirements because the Agency has the
flexibility to determine which requirements in the closure
standards are relevant and appropriate. Under this
scenario, a hybrid closure is possible. Depending on the
site circumstances and the remedy selected, clean
closure, landfill closure, or a combination of requirements
from each type of closure may be used.
The proposed revisions to the NCR discuss the concept
of hybrid closure (53 FR 51446). The NCP illustrated the
following possible hybrid closure approaches:
• Hybrid-Clean Closure. Used when leachate will not im-
pact the ground water (even though residual con-
tamination and leachate are above health-based
levels) and contamination does not pose a direct con-
tact threat. With hybrid-clean closure:
• No covers or long-term management are required.
• Fate and transport modeling and model verification
are used to ensure that ground water is usable.
• A property deed notice is used to indicate the
presence of hazardous substances.
• Hybrid-Landfill Closure. Used when residual con-
tamination poses a direct contact threat, but does not
pose a ground-water threat. With hybrid-landfill
closure:
• Covers, which may be permeable, are used to ad-
dress the direct contact threat.
• Limited long-term management includes site and
cover maintenance and minimal ground-water
monitoring.
• Institutional controls (e.g., land-use restrictions or
deed notices) are used as necessary.
The two hybrid closure alternatives are constructs of ap-
plicable laws but are not themselves promulgated at this
time. These alternatives are possible when RCRA re-
quirements are relevant and appropriate, but not when
closure requirements are applicable.
REFERENCES
1. U.S. EPA. 1989. Final covers on hazardous waste
landfills and surface impoundments. Office of Solid
Waste and Emergency Response Technical
Guidance Document EPA 530-SW-89-047, Risk
Reduction Engineering Laboratory, Cincinnati, OH.
2. U.S. EPA. 1989. RCRA ARARs: focus on closure re-
quirements. Office of Solid Waste and Emergency
Response Directive 9234.2-04FS, Office of Solid
Waste and Emergency Response, Washington, DC.
3. U.S. EPA. 1978. Closure and postclosure standards.
Draft RCRA Guidance Manual for Subpart G. EPA
530-SW-78-010. Office of Solid Waste and Emergen-
cy Response, Washington, DC.
-------
CHAPTER 2
SOILS USED IN COVER SYSTEMS
INTRODUCTION
This chapter describes several important aspects of soils
design for cover systems over waste disposal units and
site remediation projects. The chapter focuses on three
critical components of the cover system: composite ac-
tion of soil with a geomembrane liner; design and con-
struction of low hydraulic conductivity layers of
compacted soil; and mechanisms by which low hydraulic
conductivity layers can be damaged. In addition, types of
soils used for liquid drainage or gas collection also will be
discussed.
TYPICAL COVER SYSTEMS
Cover systems perform many functions. One of the prin-
cipal objectives of a cover system is to reduce leaching
of contaminants from buried wastes or contaminated
soils by minimizing water infiltration. Cover systems also
promote good surface drainage and maximize runoff. In
addition, they restrict or control gas migration, or, at
some sites, enhance gas recovery. Finally, cover sys-
tems provide a physical separation between buried
wastes or contaminated materials and animals and plant
roots. When designing a cover system, all of these re-
quirements, plus others, typically must be considered.
As presented and discussed in Chapter 1, Figures 1-1
and 1-2 illustrate two typical cover profiles (see pages 1-
3 and 1-7). Figure 1-1 illustrates the minimum cover
profile recommended by EPA for hazardous waste. Many
of the layers shown in the figure are composed of soils or
have soil components. Each layer has a different pur-
pose and the materials must be selected and the layer
designed to perform the intended function:
• Topsail - The topsoil supports vegetation (which mini-
mizes erosion and maximizes evapotranspiration),
separates the waste from the surface, stores water
that infiltrates the cover system, and protects underly-
ing materials from freezing during winter and from
desiccation during dry periods.
• Filter - The filter separates the underlying drainage
material from the topsoil so that the topsoil will not
plug the drainage material. The filter is often a geotex-
tile, but also can be soil.
• Drainage Layer - The drainage layer (which is not
needed in arid climates) serves to drain away water
that infiltrates the topsoil.
• Geomembrane Liner and Low Hydraulic Conductivity
Soil Layer- The geomembrane and low hydraulic con-
ductivity soil layer form a composite liner that serves
as a hydraulic barrier to impede water infiltration
through the cover system.
Figure 1-2 illustrates an alternative cover profile recom-
mended by EPA for hazardous waste. In Figure 1-2, cob-
bles are placed on the topsoil to provide protection from
erosion. Cobbles, which are normally used only at very
arid sites, allow precipitation to infiltrate underlying
materials, but do not promote evapotranspiration (since
there are no plants present). Figure 1-2 also depicts a
biobarrier between two filters. The biobarrier is usually a
layer of cobbles, approximately 30- to 90-cm (1- to 3-ft)
thick. The biobarrier stops animals from burrowing into
the ground, and, if the cobbles are dry, prevents the
penetration of plant roots. The gas vent layer facilitates
removal of gases that could accumulate in the waste
layer.
The cover profiles shown in Figures 1-1 and 1-2 provide
general guidance only. Depending on the specific cir-
cumstances at a particular site, some of the layers shown
in these figures may not be necessary. For example, at
an extremely arid site, a cover system placed over non-
hazardous, nonputrescible waste may simply consist of a
single layer of topsoil with no drainage layer, no hydraulic
barrier, and no gas vent layer. Conversely, some situa-
tions may require more layers than those shown in these
figures. For example, radioactive waste such as uranium
mill tailings may require a radon-emission-barrier layer. In
addition, the designer may need to include several com-
ponents or layers within the cover system to satisfy multi-
ple objectives. When such objectives lead to conflicting
technical requirements, tradeoffs are frequently neces-
sary.
FLOW RATES THROUGH LINERS
Figure 2-1 illustrates three types of hydraulic barriers
(liners) for cover systems: 1) a low hydraulic conduc-
-------
tivity, compacted soil liner; 2) a geomembrane liner; and
3) a geomembrane/soil composite liner. Flow rates for
each of these types of liners are calculated below for the
purpose of comparing the effectiveness of the barriers.
Flow rates through compacted soil liners are calculated
using Darcy's law, the basic equation used to describe
the flow of fluids through porous materials. Darcy's law
states:
q = ks i A
where q is the flow rate (m3/s); ks represents the
hydraulic conductivity of the soil (m/s); i is the dimension-
less hydraulic gradient; and A is the area (m2) over which
flow occurs. If the soil is saturated and there is no soil
suction, the hydraulic gradient (i) is:
i = (h + D) / D
where the terms are defined in Figure 2-1 (h is the depth
of liquid ponded above a liner with thickness D). For ex-
ample, if 30 cm (1 ft) of water is ponded on a 90-cm (3-ft)
thick liner that has a hydraulic conductivity of 1 x 1CT9 m/s
(1 x 10~7 cm/s), the flow rate is 120 gal (454 L)/acre/day.
If the hydraulic conductivity is increased or decreased,
the flow rate is changed proportionally (Table 2-1).
The second liner depicted in Figure 2-1 is a
geomembrane liner. It is assumed that the geomembrane
has one or more circular holes (defects) in the liner, that
the holes are sufficiently widely spaced that leakage
through each hole occurs independently from the other
holes, that the head of liquid ponded above the liner (h) is
Table 2-1. Calculated Flow Rates through Soil Liners
with 30 cm of Water Ponded on the Liner
Hydraulic Conductivity
(cm/s)
1 x10"6
1 x10'7
1 x10'8
1 x 1 O"9
Rate of Flow
(gal/acre/day)a
1,200
120
12
1
constant, and that the soil that underlies the
geomembrane has a very large hydraulic conductivity
(the subsoil offers no resistance to flow through a hole in
the geomembrane). Giroud and Bonaparte (1) recom-
mend the following equation for estimating flow rates
through holes in geomembranes under these assump-
tions:
q = CBa(2gh)05
where q is the rate of flow (m3/s); CB is a flow coefficient
with a value of approximately 0.6; a is the area (m2) of a
circular hole; g is the acceleration due to gravity (9.81
m/s2); and h is the head (m) above the liner. For ex-
ample, if there is a single hole with an area of 1 cm2
(0.0001 m2) and the head is 30 cm (1 ft) (0.305 m), the
calculated rate of flow is 3,300 gal (12,491 L)/day. If there
is one hole per acre, then the flow rate is 3,300 gal
(12,491 L)/acre/day.
Flow rates for other circumstances are calculated in
Table 2-2. Giroud and Bonaparte report that with good
quality control, one hole per acre is typical (1). With poor
control, 30 holes per acre is typical. They also note that
most defects are small (<0.1 cm2), but that larger holes
are occasionally observed. In calculating the rate of flow
for "No Holes" in Table 2-2, it was assumed that any flux
of liquid was controlled by water vapor transmission; a
Table 2-2. Calculated Flow Rates Through a
Geomembrane with a Head of 30 cm of Water
above the Geomembrane
aL = gal x 3.785
Size of Hole
(cm2)
No holes
0.1
0.1
1
1
10
Number of Holes
Per Acre
-
1
30
1
30
1
Rate of Flow
(gal/acre/day)a
0.01
330
10,000
3,300
100,000
33,000
aL = gal x 3.785
D
Area "a"
Hydraulic Conductivity "k"
o
SOIL LiNER GEOMEMBRANE
Figure 2-1. Soil liner, geomembrane liner, and composite liner.
COMPOSITE LINER
10
-------
flux of 0.01 gal/acre/day corresponds to a typical water
vapor transmission rate of geomembrane liner materials.
The third type of liner depicted in Figure 2-1 is a com-
posite liner. Giroud and Bonaparte (2) and Giroud et al.
(3) discuss seepage rates through composite liners. They
recommend the following equation for computing
seepage rates for cases in which the hydraulic seal bet-
ween the geomembrane and soil is poor:
q =1.15h°9a01 ks0.74
where all the parameters and units are as indicated pre-
viously. This equation assumes that the hydraulic
gradient through the soil is 1. If there is a good hydraulic
seal between the geomembrane liner and underlying soil,
the flow rate is approximately one-fifth the value com-
puted from the equation shown above; the constant in the
equation is 0.21 rather than 1.15 for the case of a good
seal. For example, suppose the geomembrane com-
ponent of a composite liner has one hole/acre with an
area of 1 cm2 per hole, the hydraulic conductivity of the
subsoil is 1 x 10"7 cm/s (1 x 10~9 m/s), the head of water
is 30 cm (1 ft) and a poor seal exists between the
geomembrane and soil. The calculated flow rate is 0.8
gal (3 L)/acre/day. Table 2-3 shows other calculated flow
rates for composite liners with a head of water of 30 cm
(1 ft.)
It is useful to compare the three types of liners under a
variety of assumed conditions, as illustrated in Table 2-4.
For discussion purposes, each liner type is classified as
poor, good, or excellent. EPA requires that low per-
meability compacted soil liners used for hazardous
wastes have a hydraulic conductivity no greater than 1 x
10~7 cm/s; therefore, a soil liner with a hydraulic conduc-
tivity of 1 x 10~7 cm/s is described in Table 2-4 as a
"good" liner. A compacted soil liner with a 10-fold higher
hydraulic conductivity is described as a "poor" liner, and
a soil liner with a 10-fold lower hydraulic conductivity is
described as an "excellent" liner.
For geomembrane liners, a liner with a large number of
small holes (30 holes/acre, with each hole having an area
of 0.1 cm2) is described as a "poor" liner because Giroud
and Bonaparte suggest that such a large number of
defects would be expected only with minimal construction
quality control (1). A "good" geomembrane liner was as-
sumed to have been constructed with good quality as-
surance and an "excellent" geomembrane liner was
assumed to have one small hole/acre (1). For all of the
seepage rates computed for composite liners in Table 2-
4, it was assumed that there was poor contact between
the geomembrane and soil.
As Table 2-4 illustrates, a composite liner (even one built
by poor to mediocre standards) significantly outperforms
a soil liner or a geomembrane liner alone. For this
reason, a composite liner is recommended when there is
enough rainfall to warrant a very low-permeability
hydraulic barrier in the cover system.
Table 2-3. Calculated Flow Rates for Composite Liners
with a Head of Water of 30 cm
Hydraulic
Conductivity
of Subsoil
(cm/s)
Size of Hole in
Geomembrane
(cm2)
Number of
Holes/Acre
Rate of Flow
(gal/acre/dayf
1 xicr6
1 x 1 O"6
1 x 1 O"6
1 x 1 0"6
1 xirj6
1 x 1 O"7
1 x1U7
1 x10"7
1 x 10'7
1 x 1 0"7
1 x 10'8
1 x 1 O"8
1 x 1 O"8
1 x10~8
1 x10'8
1 X 10-9
1x10'9
1 x 1 O'9
1 x 1 O"9
1x10'9
0.1
0.1
1
1
10
0.1
0.1
1
1
10
0.1
0.1
1
1
10
0.1
0.1
1
1
10
1
30
1
30
1
1
30
1
30
1
1
30
1
30
1
1
30
1
30
1
3
102
4
130
5
0.6
19
0.8
24
1.0
0.1
3
0.1
4
0.2
0.2
0.6
0.03
0.8
0.03
al_ = gal x 3.785
To maximize the effectiveness of a composite liner, the
geomembrane must be placed to achieve a good
hydraulic seal with the underlying layer of low hydraulic
conductivity soil. As shown in Figure 2-2, the composite
liner works by limiting the flow of fluid in the soil to a very
small area. Fluid must not be allowed to spread laterally
along the interface between the geomembrane and soil.
To ensure good hydraulic contact, the soil liner should be
smooth-rolled with a steel-drummed roller before the
geomembrane is placed, and the geomembrane should
have a minimum number of wrinkles when it is finally
covered. In addition, high-permeability material, such as
a sand bedding layer or geotextile, should not be placed
between the geomembrane and low hydraulic conduc-
tivity soil (Figure 2-2) because this will destroy the com-
posite action of the two materials.
If there are concerns that rocks or stones in the soil
material may punch holes in the geomembrane, the
stones should be removed, or a stone-free material with
a low hydraulic conductivity placed on the surface.
Vibratory screens also can be used to sieve stones prior
to placement. Alternatively, mechanical devices that
sieve stones or move them to a row in a loose lift of soil
may be used. A different material, or a differently
11
-------
Clav Liner
Composite Liner
Leachate
Leachate
FML
A = Area of Entire Area < Area of Entire
Liner Liner
Leachate
Do.
Leachate
Don't
Figure 2-2. Soil liner and composite liner.
processed material that has fewer and smaller stones,
may be used to construct the uppermost lift of the soil
liner (i.e., the lift that will serve as a foundation for the
geomembrane).
CRITICAL PARAMETERS FOR SOIL LINERS
Materials
The primary requirement for a soil liner material is that it
be capable of being compacted to produce a suitably low
hydraulic conductivity. To meet this requirement, the fol-
lowing conditions should be met:
• Fines - The soil should contain at least 20 percent
fines (fines are defined as the percentage, on a dry-
weight basis, of material passing the No. 200 sieve,
which has openings of 0.075 mm).
• Plasticity Index - The soil should have a plasticity
index of at least 10 percent, although some soils with a
slightly lower plasticity index may be suitable. Soils
with plasticity indices less than about 10 percent have
very little clay and usually will not produce the neces-
sary low hydraulic conductivity. Soils with plasticity in-
dices greater than 30 to 40 percent are difficult to work
with, as they form hard chunks when dry and sticky
clods when wet, which make them difficult to work with
in the field. Such soils also tend to have high
shrink/swell potential and may not be suitable for this
12
-------
Table 2-4. Calculated Flow Rates for Soil Liners,
Geomembrane Liners, and Composite Liners
Assumed Rate of
Overall Values of Flow
Type of Quality Key (gal/
Liner of Liner Parameters acre/day)3
Compacted
Soil
Geomembrane
Composite
Compacted
Soil
Poor
Poor
Poor
Good
ks=1 x1(rbcm/s 1,200
30 holes/acre;
a=0.1 cm2
10,000
ks=1 x 10~6cm/s 100
30 holes/acre;
a=0.1 cm2
-7.
Geomembrane Good
ks=1 x 10"'cm/s
1 hole/acre;
a=1 cm
Composite
Good
-7,
Compacted Excellent
Soil
Geomembrane Excellent
Composite
Excellent
ks=1 x 10~'cm/s
1 hole/acre;
a=1 cm2
k5=1 x10'8cm/s
1 hole/acre;
a=0.1 cm2
ks=1 x 10~8cm/s
1 hole/acre;
a=0.1 cm2
120
3,300
0.8
12
330
0.1
aL = gal x 3.785
reason. Soils with plasticity indices between ap-
proximately 10 and 35 percent are generally ideal.
Percentage of Gravel - The percentage of gravel
(defined as material retained on the No. 4 sieve, which
has openings of 4.76 mm) must not be excessive. A
maximum amount of 10 percent gravel is suggested as
a conservative figure. For many soils, however, larger
amounts may not necessarily be deleterious if the
gravel is uniformly distributed in the soil and does not
interfere with compaction by footed rollers. For ex-
ample, Shakoor and Cook found that the hydraulic
conductivity of a compacted, clayey soil was insensi-
tive to the amount of gravel present, as long as the
gravel content did not exceed 50 percent (4). Gravel is
only deleterious if the pores between gravel particles
are not filled with clayey soil and the gravel forms a
continuous pathway through the liner. The key
problem to be avoided is segregation of gravel in pock-
ets that contain little or no fine-grained soil.
• Stones and Rocks - No stones or rocks larger than 2.5
to 5 cm (1 to 2 in.) in diameter should be present in the
liner material.
If the soil material does not contain enough clay or other
fine-grained minerals to be capable of being compacted
to the desired low hydraulic conductivity, commercially
produced clay minerals, such as sodium bentonite, may
be mixed with the soil. Figure 2-3 shows the relationship
between the percentage of bentonite added to a soil and
the hydraulic conductivity after compaction for a well-
graded, silty soil that was carefully mixed in the
laboratory. The percentage of bentonite is defined as the
dry weight of bentonite divided by the dry weight of soil to
which the bentonite is added (Wb/Ws). For well-graded
soils containing a wide range of grain sizes, adding just a
small amount of bentonite will usually lower the hydraulic
conductivity of the soil to below 1 x 10~7. For poorly
graded soils, e.g., those with a uniform grain size, more
bentonite is often needed.
Bentonite can be added to soil in two ways. One techni-
que is to spread the soil to be amended over an area in a
loose lift approximately 23 to 30 cm (9- to 12-in.) thick.
Bentonite is then applied to the surface at a controlled
rate and mixed into the soil using mechanical mixing
equipment, such as a rototiller or road reclaimer
(recycler). Multiple passes of the mixing equipment are
usually recommended. The second procedure is to mix
the ingredients in a pugmill, which is a large device used
to mix bulk materials such as the ingredients that form
Portland cement concrete. Bulk mixing in a pugmill usual-
ly provides more controlled mixing than combining in-
gredients in place in a loose lift of soil. However, mixing
of bentonite into a loose lift of soil can be adequate if the
mixing is done carefully with multiple passes of mechani-
cal mixers and careful control over rates of application
and depth of mixing. The reason why bulk mixing is
usually recommended is that control over the mixing
process is easier.
Water Content
The water content of the soil at the time it is compacted is
an important variable controlling the engineering proper-
ties of soil liner materials. The lower half of Figure 2-4
shows a soil compaction curve. If soil samples are mixed
at several water contents and then compacted with a
consistent method and energy of compaction, the result
is the relationship between dry unit weight and molding
water content shown in the lower half of Figure 2-4. The
molding water content at which the maximum dry unit
weight is observed is termed the "optimum water content"
and is indicated in Figure 2-4 with a dashed vertical line.
Soils compacted at water contents less than optimum
("dry of optimum") tend to have a relatively high hydraulic
conductivity whereas soils compacted at water contents
greater than optimum ("wet of optimum") tend to have a
low hydraulic conductivity. It is usually preferable to com-
13
-------
E
o
>>
'>
o
"O
c
o
O
o
"B
2
•^
x
Percent Bentonite
Figure 2-3. Effect of bentonite upon the hydraulic conductivity of a bentonite-amended soil.
Wb
(Percent bentonite = TTT- )
"S
pact the soil wet of optimum to achieve minimal hydraulic
conductivity.
Figures 2-5 to 2-7 illustrate for a highly plastic soil why
wet-of-optimum compaction is so effective in achieving
low hydraulic conductivity. These three photographs
show a soil that was compacted with standard Proctor
energy (ASTM D698). The soil had a plasticity index of
41 percent. The optimum water content for this soil and
compaction procedure was 19 percent. The specimen
shown in Figure 2-5 was compacted at a water content of
12 percent (7 percent dry of optimum). This compacted
soil had a very high hydraulic conductivity (1 x 10 cm/s)
because the dry, hard clods of soil were not broken down
and remolded by the energy of compaction. The
specimen shown in Figure 2-6 was compacted at a water
content of 16 percent (3 percent dry of optimum) and had
a hydraulic conductivity of 1 x 10 cm/s; the clods were
still too dry and hard at this water content to permit the
clods to be remolded into a homogeneous mass with low
hydraulic conductivity. The specimen shown in Figure 2-7
was compacted at a water content of 20 percent (1 per-
cent wet of optimum) and had a hydraulic conductivity of
1 x 10~9 cm/s. At this water content, the clods were wet,
soft, and easily remolded into a homogeneous mass that
was free of remnant clods and large inter-clod voids and
pore spaces. The visual differences between specimens
compacted dry versus wet of optimum are usually not as
obvious as they are in Figures 2-5 to 2-7 for soils of lower
plasticity. However, even for low-plasticity clays, ex-
perience has almost always shown that the soil must be
compacted wet of optimum water content to achieve min-
imum hydraulic conductivity.
The water content of the soil must be adjusted to the
proper value prior to compaction and the water should be
uniformly distributed in the soil. If the soil requires addi-
tional water, it can be added with a water truck; care
should be taken to apply the water to the soil in a control-
led, uniform manner, e.g., with a spray bar mounted on
the rear of the trucks. Rototillers (Figure 2-8) are very ef-
fective for mixing wetted soil; these devices distribute the
water uniformly among clods of material. Figure 2-9
depicts the teeth on the blades of a rototiller, which
provide the mixing action. Mechanical mixing to mix
water evenly into the soil is especially important for highly
plastic soils that form large clods of soil.
Compactive Energy
Another important variable controlling the engineering
properties of soil liner materials is the energy of compac-
tion. As shown in Figure 2-10, increasing the energy of
compaction increases the dry unit weight of the soil,
decreases the optimum water content, and reduces
14
-------
Hydraulic
Conductivity
Dry Unit
Weight
Molding Water Content
Figure 2-4. Hydraulic conductivity and dry unit weight versus molding water content.
hydraulic conductivity. The hydraulic conductivity of a soil
that is compacted wet of optimum could be lowered by
one to two orders of magnitude by increasing the energy
of compaction, even though the dry unit weight of the soil
is not increased measurably. More energy of compaction
helps to remold clods of soil, realign soil particles, reduce
the size or degree of connection of the largest pores in
the soil, and lower hydraulic conductivity.
The compactive energy delivered to soil depends on the
weight of the roller, the number of passes of the roller
over a given area, and the thickness of the soil lift being
compacted. Increasing the weight and number of passes,
and decreasing the lift thickness, can increase the com-
pactive effort. The best combination of these factors to
use when compacting low hydraulic conductivity soil
liners depends on the water content of the soil and the
firmness of the subbase.
Heavy rollers cannot be used if the soil is very wet or if
the foundation is weak and compressible (e.g., if
municipal solid waste is located just 30- to 60-cm [1 - to 2-
ft] below the layer to be compacted). Rollers with static
weights of at least 13,608 to 18,144 kg (30,000 to 40,000
pounds) are recommended for compacting low hydraulic
conductivity layers in cover systems. Rollers that weigh
up to 31,752 kg (70,000 pounds) are available and may
be desirable for compacting bottom liners of landfills, but
such rollers are too heavy for many cover systems be-
cause of the presence of compressible waste material a
short distance below the cover.
The roller must make a sufficient number of passes over
a given area to ensure adequate compaction. The mini-
mum number of passes will vary, but at least 5 to 10 pas-
ses are usually required to deliver sufficient compactive
energy and to provide adequate coverage.
15
-------
o
, * A f 3', i A Hi.
Pi. "X * >R
Figure 2-5. Highly plastic soil compacted with standard
Proctor procedures at a water content of 12%.
16 %
STANDARD
PROCTOR
Figure 2-6. Highly plastic soil compacted with standard
Proctor procedures at a water content of 16%.
STANDARD
PROCTOR
Figure 2-7. Highly plastic soil compacted with standard
Proctor procedures at a water content of 20%.
Size of Clods
The clay-rich soils that are usually used to construct soil
liners typically form dry, hard clods of soil or wet, sticky
clods, depending on water content. Highly plastic soils al-
most always form large clods. Soils with low plasticity
(plasticity index less than about 10%) do not form very
large clods. For soils that form clods, the clods must be
remolded into a homogeneous mass that is free of large
inter-clod pores if low hydraulic conductivity is to be
achieved.
Benson and Daniel described the influence of clod size of
a highly plastic soil (plasticity index = 41%) upon
hydraulic conductivity (5). These investigators processed
a clayey soil by breaking clods down to pass either the
No. 4 sieve (4.76 mm or 0.2 in. openings) or the 1.9-cm
(3/4-in.) sieve. The soil was then wetted, allowed to
hydrate at least 24 hours, compacted, and permeated.
Benson and Daniel's (1990) results are summarized in
Table 2-5. The optimum water content was 17 percent for
the clods processed through the sieve with a 0.5-cm (0.2-
in.) opening and 19 percent for the soil processed
through the sieve with a 1.9-cm (3/4-in.) opening. For soil
compacted dry of optimum, the soil with smaller clods
had a hydraulic conductivity that was several orders of
magnitude lower than the soil with larger clods. When the
soils were compacted wet of optimum, the size of clods
had a negligible effect. Size is therefore important for dry,
16
-------
Figure 2-8. Rototiller used to mix soil.
Figure 2-9. Blades and teeth on rototiller.
17
-------
Table 2-5. Effect of Size of Clods during Processing of
Soil upon Hydraulic Conductivity of Soil after
Compaction
Hydraulic Conductivity (cm/s)
Molding Water
Content (%) 0.2-in. Clods3 0.75-in. Clods3
12 2x10"8
16 2 x 1 0"9
18 1 x 1 0"8
20 2x10'9
4x 10'4
1 x10'3
8x10-10
7x10-10
acm = in. x 2.540
hard clods (dry of optimum), but not for wet, soft clods
(wet of optimum). When the soil is compacted wet of op-
timum, the clods are sufficiently soft that they are easily
remolded regardless of theiroriginal size.
One way to reduce the size of clods in dry materials is to
use a road reclaimer (also called a road recycler), such
as the one shown in Figure 2-11. This device pulverizes
materials with teeth that rotate on a drum at a high
speed. The device was used with great effectiveness at a
site in Pennsylvania in which a mudstone was used for a
liner material (Figure 2-12). In the figure, the road
reclaimer has made a pass through a loose lift of
material. After just one pass of the road reclaimer, the
size of mudstone clods has been greatly reduced.
Bonding of Lifts
Bonding of lifts is important in achieving a low hydraulic
conductivity in soil liners. The upper half of Figure 2-13 il-
lustrates a cross-section of a soil liner consisting of four
lifts. A borehole has been drilled into the lowest lift, filled
with a dye-stained fluid, left for a period of time, and then
drained. The dye penetrates the soil further along lift in-
terfaces than through the lifts themselves. Due to imper-
fect bonding of lifts, a zone of higher horizontal hydraulic
conductivity exists at lift interfaces in this example.
Lift interfaces have important ramifications with respect to
the overall hydraulic performance of a soil liner. The
lower half of Figure 2-13 depicts a liner consisting of six
lifts. Each lift has a few "hydraulic defects." If the lift in-
terfaces have high hydraulic conductivity, water can flow
downward through the more permeable zones in a lift
and spread laterally along a lift interface until it en-
counters a permeable zone in the underlying lift. This
process repeats for underlying lifts and lift interfaces. In
this way lift interfaces provide hydraulic connection bet-
ween defects in overlying and underlying lifts. Better
overall performance (lower hydraulic conductivity) is
achieved if lifts are bonded together to eliminate high
conductivity at lift interfaces.
To bond lifts together, the surface of the previously
compacted lift should be rough so that the newly
placed lift can effectively blend into the surface. If
necessary, the surface of the previously compacted lift
can be roughened by discing the soil to a depth of ap-
proximately 2.5 cm (1 in). Discing the soil involves
plowing up the soil surface to a shallow depth so that
the surface is rough and so that there will be no abrupt
interface between lifts.
Compactors with long "feet" on the drums are useful in
blending one lift into another. Figure 2-14 shows a
popular heavy compactor (20,000 kg [44,000 pounds])
with feet that are 18 to 23 cm (7 to 9 in.) long. During the
first few passes of the compactor, the feet sink through a
loose lift of soil and compact the newly placed lift into the
surface of the previously compacted lift. Using a roller
with feet that fully penetrate a loose lift of soil is recom-
mended to bond lifts and to minimize high horizontal
hydraulic conductivity at lift interfaces.
If a geomembrane liner will be placed on the compacted
soil liner, the final surface of the soil liner should be com-
pacted with a smooth, steel drum roller to achieve a good
hydraulic seal.
EFFECTS OF DESICCATION
Desiccation of soil liners occurs whenever the soil liner
dries, which can be during or after construction. Desicca-
tion causes soil liner materials to shrink and, potentially,
to crack. Cracking can be disastrous in terms of hydraulic
conductivity because cracked liners are more permeable
than uncracked liners.
Boynton and Daniel desiccated slabs of compacted clay,
trimmed cylindrical test specimens for hydraulic conduc-
tivity testing from the desiccated slabs, and measured the
hydraulic conductivity at different effective confining
stresses (6). In laboratory tests, the confining stress
simulates the weight of overburden soil; the greater the
confining stress, the greater the depth of burial below the
surface that is simulated. Control tests also were per-
formed on soils that had not been desiccated. These
results are summarized in Figure 2-15. At low confining
stress, the desiccated soils were much more permeable
than the control. At high confining stress, however, the
desiccated soils were no more permeable than the con-
trol. It appeared that the application of a large compres-
sive stress (>5 psi, or 35 kPa) closed the desiccation
cracks that had formed and, in combination with hydra-
tion of the soil, essentially fully healed the damage done
by desiccation.
In cover systems, the overburden stress on the liner com-
ponents is controlled by the depth of soil overlying the
liner. Because the thickness of soil overburden above the
liner seldom exceeds a few feet, the overburden stress is
normally low. Soil applies an overburden stress of ap-
proximately 1 psi per foot of depth. Thus, for example, if
18
-------
o
13
~o
C^^^^^
f ^\
O E
=3
05
D)
'CD
o
10
-5
10
-6
-7
10
,o8
116
108
100
92
1 1 1 1 1 r
Increasing
Compactive
Effort
Optimum
Water
Content
Increasing Compactive
Effort
i i i i
12 14 16 18 20 22 24
Molding Water Content (%)
Figure 2-10. Influence of Compactive effort upon hydraulic conductivity and dry unit weight.
60 cm (2 ft) of topsoil overlie a 60-cm (2-ft) thick layer of
compacted clay, the maximum overburden stress at the
bottom of the clay is approximately 4 psi. Based on Boyn-
ton and Daniel's results, if desiccation of the compacted
soil liner occurs in a cover system, even though wetting
of the soil may partly swell the soil and "heal" desiccation
cracks, it is not expected that all the damage done by
desiccation would be self-healing.
Montgomery and Parsons described an example of the
damaging effects of desiccation (7). Test plots were built
at the Omega Hills Landfill near Milwaukee, Wisconsin, in
1985. In both test plots, the cover systems consisted of
122 cm (4 ft) of compacted clay. The clay was overlain by
15 cm (6 in.) of topsoil in one plot and 46 cm (18 in.) of
topsoil in the other. In both test plots, the upper 20 to 25
cm (8 to 10 in.) of compacted clay had weathered and
become blocky after 3 years. Cracks up to 1.3-cm (1/2-
in.) wide extended 89 to 102 cm (35 to 40 in.) into the
compacted clay liner. The 46 cm (18 in.) of topsoil did not
appear to be any more effective than 15 cm (6 in.) in
protecting the underlying clay from desiccation.
The layer of low hydraulic conductivity, compacted soil
placed in a cover system must be protected from the
damaging effects of desiccation both during and after
19
-------
Figure 2-11. Road recycler used to pulverize clods of soil.
construction. During construction, the soil must not be al-
lowed to dry significantly either during or after compac-
tion of each lift. Frequent watering of the soil is usually
the best way to prevent desiccation during construction.
The higher the water content of the soil and the higher
the plasticity of the soil, the greater is the shrinkage
potential from desiccation. There are two ways to provide
the required protection after construction. One way is to
bury the liner beneath an adequate depth of soil overbur-
den; another technique is to place a geomembrane over
the soil. If a geomembrane liner is placed on a soil liner
to form a composite, it is often convenient to overbuild
the soil liner (i.e., make it thicker than necessary) and
then to scrape away a few inches of potentially desic-
cated surficial soil just before the geomembrane is
placed.
EFFECTS OF FROST ACTION
Zimmie and La Plante studied the effects of freezing and
thawing upon the hydraulic conductivity of a compacted
clay by testing soils compacted dry of optimum, at op-
timum, and wet of optimum (8). They found that
freeze/thaw cycles caused an increase in hydraulic con-
ductivity of one to two orders of magnitude in all soils ex-
amined. Most of the damage was done after only one to
two cycles of freezing and thawing. From this and other
work, it is recommended that the low hydraulic conduc-
tivity component of cover systems not be allowed to
freeze. Freezing can be avoided by burying the low
hydraulic conductivity soil layer under an adequately thick
layer of soil.
EFFECTS OF SETTLEMENT
Two types of settlement are of concern with respect to
covers: total settlement and differential settlement. Total
settlement of the surface of a cover is the total downward
movement of a fixed point on the surface. Differential set-
tlement is always measured between two points and is
defined as the difference between the total settlements at
these two points. Distortion is defined as the differential
settlement between two points divided by the distance
along the ground surface between the two points. Exces-
sive differential settlement of underlying waste can
damage a cover system, if differential settlement occurs,
tensile strains develop in cover materials as a result of
bending stresses and/or elongation. Tensile strain is
defined as the amount of stretching of an element divided
by the original length of the element. Anytime the cover
settles differentially, some part of the cover will be sub-
jected to tension and will undergo tensile strain. Tensile
strains are of concern because the larger the stretching
(tensile strain), the greater the possibility that the soil will
crack and that a geomembrane will rupture. Bending
stresses, stresses that occur when an object is bent,
result when covers settle differentially; part of the bent
cover is in tension and part is in compression. Bending
stresses are of concern because the tensile stresses as-
sociated with bending may be large enough to cause the
20
-------
'Sjii*^ •
.. .
o,-«" •', > ^»*w» .
*'"»***,>? ' " ',-,
" '' • '*;
. ,
** . *
v
'' * V,
**•»-
Figure 2-12. Passage of road recycler over loose lift of mudstone to reduce size of chunks of mudstone.
Borehole
Lift 1
Lift 2
Lift 3
Lift 4
J > C f
f c ^
^
r "\
~) } J ^^^
A r>
s r
^ C ^ )
^ r (
Figure 2-13. Effect of Imperfect bonding of lifts on hydraulic performance of soil liner.
21
-------
Figure 2-14. Example of heavy footed roller with long feet.
soil to crack. Geomembranes can generally withstand far
larger tensile strains without failing than soils. The
geomembrane also has the ability to elongate (stretch) a
great deal without rupturing or breaking.
Gilbert and Murphy discuss the prediction and mitigation
of subsidence damage to covers (9). Gilbert and Murphy
developed a relationship between tensile strain in a cover
and distortion, delta/L, where delta is the amount of dif-
ferential settlement that occurs between two points that
are a distance (L) apart. This relationship is shown in
Figure 2-16. As the distortion increases, the tensile strain
in the cover soils increases.
Minor cracking of topsoil or drainage layers as a result of
tensile stresses is of little concern. However, cracking of
a hydraulic barrier, such as a layer of low hydraulic con-
ductivity soil, is of great concern because the hydraulic
integrity of the barrier layer is compromised if it is crack-
ed. The amount of strain that a low hydraulic conductivity,
compacted soil can withstand prior to cracking depends
significantly upon the water content of the soil. As shown
in Figure 2-17, soils compacted wet of optimum are more
ductile than soils compacted dry of optimum. For cover
systems, ductile soils that can withstand significant strain
without cracking are preferred. For this reason, as well as
the hydraulic conductivity considerations discussed ear-
lier, it is preferable to compact low hydraulic conductivity
soil layers wet of optimum. The soil must then be kept
from drying out and cracking, as discussed earlier.
Gilbert and Murphy summarize information concerning
tensile strain at failure for compacted, clayey soils (9).
The available data show that such soils can withstand
maximum tensile strains of 0.1 to 1 percent. If the lower
limit (0.1 percent) is used for design, the maximum allow-
able value of distortion (delta/L) is approximately 0.05
(Figure 2-17).
To put this in perspective, suppose that a circular depres-
sion develops in a cover system. The depression has a
radius of 3 m (10 ft) (diameter=6 m [20 ft]). The maximum
allowable delta/L is 0.05, and L is the radius of the
depression, which is 3 m (10 ft). The maximum allowable
settlement (delta) is 0.05 times 3 m (10 ft), or 15 cm (6
in.). If the settlement at the center of the depression ex-
ceeds 6 in., the clay layer may crack from the tensile
strains caused by the settlement.
Some wastes (such as loose municipal solid waste or un-
consolidated sludge of varying thickness) are so com-
pressible that constructing a cover system above the
waste will almost certainly produce distortions that are far
larger than 0.05. The hydraulic integrity of a low hydraulic
conductivity layer of compacted soil is likely to be
seriously damaged by the distortion caused by large dif-
ferential settlement. If the waste is continuing to settle,
e.g., as a result of decomposition, it may be prudent to
place a temporary cover on the waste and wait for settle-
ment to take place prior to constructing the final cover
system. Alternatives for stabilizing the waste include
22
-------
~ 10
-7
0
(kPa)
50
100
o
0)
E
o
o
3
T3
c
o
o
10
-8
o
k_
•o
X
10
-9
1 ~
So mple Con t alnln g
Desiccation
Cracks
Sample Containing
No Desiccation
Cracks
8
12
16
Ef f ective Confin ing Pressure (psi)
Figure 2-15. Effect of desiccation upon the hydraulic conductivity of compacted clay (6).
deep dynamic compaction, soil preloading, and the use
of wick drains to consolidate sludges. These technologies
for waste stabilization are presently emerging and ap-
propriate descriptions are not available in the literature.
INTERRACIAL SHEAR
The stability of a cover system is controlled by the slope
angle and the friction angles between the various inter-
faces of the cover system components. One potential
problem with covers installed with a sloping surface is the
risk that all or part of the cover system may slide
downhill. The recent failure of a partly completed hazard-
ous waste landfill provides an example of the problem
(10). At this facility, slippage occurred between two com-
ponents of the liner system in the landfill cell. The cell
was filled such that a slope was created on the liner sys-
tem that caused slippage.
The interfacial shearing characteristics of all components
of a cover system, as well as internal shearing
parameters of all soil layers, must be known in order to
evaluate stability. If the soils are fully saturated and
below the free water surface, e.g., during a heavy
rainstorm, the stability is much less than if the soils are
dry. Thus, one must consider both typical and worst-case
23
-------
1.0
0.8
0.6
0.4
0.2
0.0
.1
10
100
Tensile Strain (%)
Figure 2-16. Relationship between distortion and tensile strain (9).
conditions when analyzing the stability of the cover sys-
tem.
Methods of measuring interfacial friction between
geosynthetic/geosynthetic or geosynthetic/soil interfaces
are reviewed in detail by Takasumi et al. (11). No stand-
ard testing method exists, although one is under develop-
ment by ASTM.
Seed and Boulanger (12) measured interfacial friction
angles between a smooth high density polyethylene
(HOPE) geomembrane and a compacted soil-bentonite
mixture that contained 5 percent bentonite by dry weight.
Interfacial friction angles were found to be very sensitive
to compaction water content, dry unit weight, and the
degree of wetting of the soil. For a given dry unit weight,
increasing the molding water content or wetting the com-
pacted soil reduced the interfacial friction angle. Increas-
ing the density typically reduced the interfacial friction
angle, as well. Unfortunately, the compaction conditions
that would yield minimal hydraulic conductivity (i.e., com-
paction wet of optimum with a high energy of compac-
tion) also yielded the lowest interfacial friction angles.
Seed and Boulanger reported interfacial friction angles
that were typically 5 to 10 degrees for the water
content—unit weight combinations that would typically be
employed to achieve minimal hydraulic conductivity.
The study of interfacial friction problems is an area of ac-
tive research. At the present time, designers are cau-
tioned to give careful consideration to the problem and to
measure friction angles along all potential sliding sur-
faces using the proposed construction materials for test-
ing. If adequate stability is not provided, the designer will
need to consider alternative materials (e.g., rougher
geomembranes with higher interfacial friction angles),
flatter slopes, or reinforcement of the cover, e.g., with
geogrids.
DRAINAGE LAYERS
Drainage layers are high-permeability materials used to
drain fluids (such as infiltrating water) or gas produced
from the waste. A drainage layer installed to drain in-
filtrating water is called a surface water collection and
removal system. The hydraulic conductivity required for
this layer depends upon the rate of infiltration, the slope
of the layer, and the hydraulic conductivity of the underly-
ing barrier layer. However, the efficiency of the drainage
layer improves as the hydraulic conductivity of the
drainage material increases. Thus, high hydraulic con-
ductivity is a requirement for drainage layers.
The single most important factor controlling the hydraulic
conductivity of sands and gravels is the amount of fine-
grained material present. Geotechnical engineers define
fine-grained materials as those materials that will pass
through the openings of a No. 200 sieve (0.075 mm
openings). A relatively small shift in the amount of fines
24
-------
A
c
0
Q
Q
t
C/D
05
0
Water Content
Strain
Figure 2-17. Relationship between shearing characteristics of compacted soils and conditions of compaction.
present in the soil can change the hydraulic conductivity
by several orders of magnitude. The drainage material
should be relatively free of fines if the material is to have
a high hydraulic conductivity.
A minimal amount of compaction of the drainage
materials in a cover is adequate to guard against settle-
ment; excessive compaction is usually not necessary. In
fact, excessive compaction may grind up soil particles,
which would tend to lower the hydraulic conductivity of
the drainage layer. However, sands may bulk if placed in
a damp or wet condition, which can lead to an unaccep-
tably loose material. If significant seismic ground shaking
is possible at a site, compaction of drainage layers may
be needed to minimize the risk of liquefaction-induced
sliding.
Designers often place a highly permeable layer at the
base of a cover system above gas-producing wastes,
such as municipal solid waste. This layer aids in collect-
ing gas and is called a gas collection layer. Adequate fil-
ters above and below the gas collection layer must be
provided so that the collection layer does not become
clogged with fine material. Vent pipes are normally
placed in the gas collection layer at a frequency of ap-
proximately one per acre.
SUMMARY
Soils are used in cover systems to support growth of
vegetation, to separate buried wastes or contaminated
soils from the surface, to minimize the infiltration of water,
and to aid in collecting and removing gases. The most
challenging aspect of utilizing soils in cover systems is
designing, constructing, and maintaining a barrier layer of
low hydraulic conductivity. Soils can be compacted to
achieve a low initial hydraulic conductivity, but the soils
can be damaged by excessive differential settlement,
desiccation, and other environmental stresses. Protecting
a compacted soil liner from damage is therefore the
greatest challenge to the designer.
REFERENCES
1. Giroud, J.P. and R. Bonaparte. 1989a. Leakage
through liners constructed with geomembranes—Part I.
Geomembrane liners. Geotextiles and
Geomembranes. Vol. 8: 27-67.
25
-------
2. Giroud, J.P. and R. Bonaparte. 1989b. Leakage
through liners constructed with geomembranes—Part II.
Composite liners. Geotextiles and Geomembranes.
Vol.8:71-111.
3. Giroud, J.P., A. Khatami, and K. Badu-Tweneboah.
1989. Evaluation of the rate of leakage through com-
posite liners. Geotextiles and Geomembranes. Vol. 8:
337-340.
4. Shakoor, A. and B.D. Cook. 1990. The effect of stone
content, size, and shape on the engineering proper-
ties of a compacted silty clay. Bulletin of the Associa-
tion of Engineering Geologists. Vol. 27, No. 2:
245-253.
5. Benson, C.H. and D.E. Daniel. 1990. Influence of
clods on hydraulic conductivity of compacted clay.
Journal of Geotechnical Engineering. Vol. 116, No. 8:
1231-1249.
6. Boynton, S.S. and D.E. Daniel. 1985. Hydraulic con-
ductivity tests on compacted clay. Journal of
Geotechnical Engineering. Vol. 111, No. 4: 465-478.
7. Montgomery, R.J. and L.J. Parsons. 1989. The
Omega Hills Final Cover Test Plot Study: Three-
Year Data Summary. Presented at the 1989 Annual
Meeting of the National Solid Waste Management
Association, Washington, DC.
8. Zimmie, T.F. and C. La Plante. 1990. The effect of
freeze-thaw cycles on the permeability of a fine-
grained soil. Proceedings, 22nd Mid-Atlantic
Industrial Waste Conference. Philadelphia, Pennsyl-
vania: Drexel University.
9. Gilbert, P.A. and W.L. Murphy. 1987. Predic-
tion/mitigation of subsidence damage to hazardous
waste landfill covers. EPA/600/2-87/025 (PB87-
175386). Cincinnati, Ohio: U.S. EPA.
10. Seed, R.B., J.K. Mitchell, and H.B. Seed. 1990. Ket-
tlemam Hills waste landfill slope failure. II: Stability
analyses. Journal of Geotechnical Engineering. Vol.
116, No. 4: 669-691.
11. Takasumi, D.L., Green, K.R., and R.D. Holtz. 1991.
Soil-Geosynthetics Interface Strength Charac-
teristics: A Review of State-of-the-Art Testing Proce-
dures. Proceedings, Geosynthetics 91, Vol. 1,
87-100.
12. Seed, R.B., and R.W. Boulanger. 1991. Smooth
HOPE—Clay Interface Shear Strengths: Compaction
Effects. Journal of Geotechnical Engineering. Vol.
117, No. 4, 686-693.
26
-------
CHAPTER 3
GEOSYNTHETIC DESIGN FOR LANDFILL COVERS
GENERAL COMMENTS ON
DESIGN-BY-FUNCTION
Geosynthetics (GS), like all engineering materials, are
capable of being evaluated using a design-by-function
approach which includes a traditional factor-of-safety
(FS) concept. The primary function of the geosynthetic
depends upon its location in the facility. The usual re-
quirements for waste containment systems are given in
Table 3-1.
Table 3-1. Customary Primary Functions of Geosyn-
thetics Used in Waste Containment Systems
Primary Function
Type of
Geosynthetic Separate Reinforce Filter Drain Barrier
Geomembrane •
Geotextile • • • •
Geonet •
(Geo) pipe •
Geocomposite •
Geogrid •
Upon selection of the primary function, the FS should be
calculated in the same manner as for any other engineer-
ing material:
Allowable (Test) Value
FS =
Required (Design) Value
(D
In the above equation, the test values usually come from
ASTM Committee D35 on Geosynthetics, or some other
standardization group. The design values come from a
site-specific situation that utilizes relevant aspects of
geotechnical, hydraulic, polymer, or environmental en-
gineering principles or from the appropriate governing
regulations. The actual magnitude of the FS is a reflec-
tion of the degree of certainty of the design as well as the
implications of the system's nonperformance.
GEOMEMBRANE DESIGN CONCEPTS
For the design of a geomembrane in a landfill cover there
are at least four general considerations: liner com-
patibility, vapor transmission (water and methane),
biaxial stresses via subsidence, and planar stresses mo-
bilized by friction. Each will be described separately
below.
Geomembrane Compatibility
Since the liquid interfacing the geomembrane liner is
generally water, there is usually no need for EPA 9090
chemical compatibility testing, except in unusual cir-
cumstances. Some other tests that might be considered,
however, are the following:
• Dimensional stability test via ASTM D-1204
• Resistance to soil burial via ASTM D-3083
• Water extraction test via ASTM D-3083
• Volatile loss test via ASTM D-1203
• Biological resistance test via ASTM G-22
• Fungus resistance test via ASTM G-21
The use of these tests is required on a site-specific and
geomembrane-specific basis.
Vapor Transmission
Testing of water vapor transmission through geomembranes
is performed via ASTM E-96. The EPA technical resources
document of September 1988 (1) gives the following values
for the indicated geosynthetic materials:
PVC (polyvinyl chloride)
- 30 mil -1.9 g/m2-day
CPE (chlorinated polyethylene)
- 40 mil - 0.4 g/m2-day
CSPE (chlorylsulfonated polyethylene)
- 40 mil - 0.4 g/m2-day
HOPE (high density polyethylene)
- 30 mil - 0.02 g/m2-day
HOPE - 98 mil - 0.006 g/m2-day
The conversion from g/m2-day to gal/acre-day is 1 to
1.07.
A related measurement is methane (ChU) gas transmis-
sion through geomembranes. This lighter-than-air gas will
rise up from the waste and interface with the
geomembrane. Methane gas transmission rates for dif-
27
-------
ferent geomembranes have been reported in EPA's tech-
nical resources document as follows:
Yes
Hcs
PVC
-10 mil - 4.4 ml/m -day-atm.
- 20 mil - 3.3 ml/m -day-atm.
LLDPE (linear low density polyethylene)
-18 mil - 2.3 ml/m2-day-atm.
CSPE - 32 mil - 0.27 ml/m2-day-atm.
- 34 mil -1.6 ml/m2-day-atm.
HOPE - 24 mil -1.3 ml/m2-day-atm.
- 34 mil -1.4 ml/m2-day-atm.
Biaxial Stresses via Subsidence
As the waste beneath the closure subsides, differential
settlement is likely to occur. Thus a factor-of-safety for-
mulation of FS = aaiiow/Oreqd is necessary. This situation
has been modeled (see Appendix A, Stability and Ten-
sion Considerations Regarding Cover Soils on
Geomembrane Lined Slopes), giving rise to the following
formula for required strength (areqd)):
where
CFreqd -
Yes
Hcs
t
D, L
2 D L* Yes Hcs
3 t (D2 + L2)
unit weight of cover soil
height of cover soil
thickness of geomembrane
see Figure 3-1
The allowable strength aaiiow of the candidate
geomembrane must be evaluated in a closely simulated
test, e.g., GRI's GM-4 entitled "Three Dimensional
Geomembrane Tension Test." Figure 3-2 presents the
response to this test of a number of common
geomembranes used in closure situations.
Planar Stresses via Friction
In addition to the above out-of-plane stresses, the cover
soil over the geomembrane might develop greater fric-
tional stresses than the soil material beneath it. This hap-
pens particularly if a wet-of-optimum clay is placed
beneath. Again a factor-of-safety formulation is formed by
comparing the allowable strength (Taiiow) to the required
strength (Treqd) but now in force units rather than stress
units, e.g., FS = Taiiow/Treqd. The required geomembrane
tension can be obtained by the equation given in Figure
3-3 (see Appendix A for a more detailed discussion).
where cau, C
8u,5i_
(0
L
W
adhesion of the material upper
and lower of the geomembrane
friction angle of the material
upper and lower of the
geomembrane
slope angle
slope length
unit width of slope
unit weight of cover soil
height of cover soil
When calculated, the value of Treqd in Figure 3-3 is com-
pared to the Taiiow of the candidate geomembrane. This
value is currently taken from ASTM D-4885, the wide-
width tensile test for geomembranes. Note that this value
must be suitably adjusted for creep, long-term degrada-
tion, and any other site-specific situations that are con-
sidered relevant.
GEONET AND GEOCOMPOSITE SHEET DRAIN
DESIGN CONCEPTS
Geonets and/or geocomposite drains are often used as
surface water drains located immediately above the
geomembrane in a landfill closure system. There are
three aspects to the design that require attention:
material compatibility, crush strength, and flow capability.
Compatibility
Since the liquid being conveyed by the geonet or
drainage geocomposite is water, EPA 9090 testing is
usually not warranted. The polymers from which these
products are made are polyethylene (PE), polypropylene
(PP), high-impact styrene (HIS) or other long-chain
molecular structures that have good water resistance and
long-term durability when covered by soil.
Crush Strength
The crush strength of the candidate product must be
evaluated by comparing an allowable strength to a re-
quired stress, i.e., FS = aaiiow/o>eqd. The allowable
strength is taken as the rib lay-down for geonets and the
telescoping crush strength for drainage geocomposites.
Figure 3-4 illustrates common behavior for geonets and
geocomposites. The test methods currently recom-
mended are GRI GN-1 for compression behavior of
geonets and GRI GC-4 for drainage geocomposites, i.e.,
for sheet drains.
The required stress is the dead load of the cover soil plus
any live loads that may be imposed, such as construction
and maintenance equipment.
Figure 3-1. Required strength.
28
-------
5000
co
20
40 60
STRAIN (%)
80
1 00
Figure 3-2. Response of common geomembranes to the three-dimensional geomembrane tension test.
req'd
ca,, ca, ) + Y.O H,,o cos w(tan Sy - tan 5LJ| L . W
reqd [\. u \.j •
-------
Stress
Geocomposite
Strain
Figure 3-4. Common crush strength behavior for geonets and geocomposites.
GEOPIPE AND GEOCOMPOSITE EDGE DRAIN
DESIGN CONCEPTS
All of the infiltrating surface water that is collected from
the geonet or geocomposite sheet drain is conveyed to
the perimeter of the closure, where it is collected in a per-
forated pipe or in a geocomposite edge drain. The design
of the geopipe or geocomposite edge drain must con-
sider compatibility, crush strength, and flow rate.
Compatibility
With surface water as the flow medium, EPA 9090
chemical resistance testing is usually not required. Plas-
tic pipe is usually unplasticized PVC or HOPE, and al-
most all edge drains are HOPE. Thus, resistance to water
is very good.
Crush Strength
The formulation for crush strength is straightforward, FS
= aaiiow/areqd. The allowable strength for plastic pipe is
ASTM D-2412 and for geocomposite edge drain cores is
GRI's GC-4. The response to both types of materials is
very crisp (see Figure 3-6).
The required strength is the dead weight of the cover soil
plus any live loads that might be superimposed onto the
system. Truck traffic around the edge of the closure
should be considered in this regard.
Flow Rate
The required flow rate of the pipe or edge drain is the
cumulative flow coming from the geonet or sheet drain
above the geomembrane. Furthermore, this flow is again
cumulative within the pipe or edge drain and is at its max-
imum at the drainage outlet. Required flow rate will
probably dictate the frequency and location of outlets.
To determine the allowable flow rate of pipes, the Man-
ning formula is usually used with a roughness coefficient
for smooth plastic pipe of 0.015. This is straightforward
hydraulics engineering for pipes flowing partially full. For
geocomposite edge drains, the ASTM D-4716 test
described earlier is recommended.
GEOTEXTILE FILTER DESIGN
CONSIDERATIONS
Geotextiles have the greatest flexibility to serve a number
of different functions. In a landfill closure, one of the most
important is as a filter allowing water to enter a drainage
material composed of stone, geonet, geocomposite, or
perforated pipe. In selecting a geolextile for this purpose,
compatibility, hydraulic conductivity, soil retention, and
long-term clogging all must be addressed.
Compatibility
The vast majority of geogrids are PP or PET. Both of
these polymers are very stable in contact with water and
have demonstrated good durability properties. Generally,
there is no need to perform an EPA 9090 chemical com-
patibility test. The geotextile literature is abundant with
related tests assessing long-term behavior and perfor-
mance (2).
Permeability
A geotextile filter must have sufficient openings to allow
the water to enter the drain without developing excess
pore water pressure in the upstream soil. Such pressures
could mobilize cover soil instability. The flow rate design
is based on permittivity, \|/ = kn/t, where kn = hydraulic
conductivity normal to the fabric, and t = the fabric thick-
ness. As usual, a FS = vaiiowAj'reqd is formulated. The al-
lowable value is obtained from ASTM D-4491 but the
30
-------
ASTM O-4716 Flow Rate Test
Constant Load
Overflow
5000
10OOO 15000
?* fib ft.1!
20000
Figure 3-5. ASTM D-4716 flow rate test.
31
-------
Igallow
Stress
Strain
Figure 3-6. Crush strength of geopipe and geocomposite edge drain cores.
result must be modified to site-specific conditions using
partial factors of safety. The required value comes from a
water balance method, e.g., the HELP computer code
(see Chapter 8).
Geotextile Soil Retention
The geotextile filter must retain the upstream cover soil,
thereby requiring the voids to be suitably small. Since the
desired retention is the opposite of permittivity, the design
must balance the two conflicting design considerations.
Fortunately, there are about 4,000 commercially available
geotextiles to choose from, thus a design can generally
be satisfied by a number of products. The factor-of-safety
formulation is as follows:
FS=
where
des
095
2 to 4
the 95 percent finer soil size
the geotextile's 95 percent
opening size
The former value (des) is obtained by dry sieving the
upstream soil, the latter value is obtained by sieving glass
beads through the fabric as per ASTM D-4751.
Geotextile Clogging Evaluation
There are four laboratory tests to evaluate long-term
geotextile clogging by upstream soil particles:
• Gradient ratio test (GR).
• Long-term flow test (LTF).
• Hydraulic conductivity ratio test (HCR).
• Fine fraction filtration test (F3).
The first two tests are established in the literature and are
recommended for geotextiles used in landfill covers. The
latter two tests are experimental and aimed at situations
of relatively severe clogging. In general, use of geotextile
filters in landfill closures as discussed in this chapter is
not a particularly demanding design situation.
GEOGRID, OR GEOTEXTILE, COVER SOIL
REINFORCEMENT
Due to the relatively low interface friction angle of many
geomembranes, there have been some instances of
cover soils sliding off the liner in a gradual (or sometime
abrupt) manner. Many of these situations can be avoided
using geogrid, or geotextile reinforcement. The procedure
is sometimes called "veneer reinforcement." The
design is based on a factor-of-safety formulation of FS=
Taiiow/Treqd. The allowable strength is obtained from a
wide-width tensile strength test such as ASTM D-4595.
This value, however, must be reduced for such site-
specific conditions as installation damage, creep, and
long-term degradation.
Based on an infinite slope analysis, the required strength
of the geogrid (or geotextile) is given by the equation in
Figure 3-7. Note that the analysis includes a seepage
force "S" and an earthquake force "E." Both are, of
course, site specific and, if large, can easily dominate the
design.
GEOTEXTILE METHANE GAS VENT
Beneath the liner system in a landfill closure, gases
lighter than air will accumulate and gradually exert pres-
sure on the underside of the geomembrane. In some
known cases, four feet of cover soil have been cast off of
32
-------
T [Required Geogrid (or GT) Strength]
Tension
Crack
SQJ|
Weight Y)
Passive Wedge
L (Slope Length)
-/I
- 1
IH
LH "
2H2
\
r
. 5 )
sin
cos (
where
S = Possible Seepage Force
E = Possible Earthquake Force
Figure 3-7. Required strength of geogrid for cover soil reinforcement.
the geomembrane and a geomembrane "whale" has ap-
peared above the ground surface. To avoid such occur-
rences, these landfill gases (mainly methane) must be
conveyed along the underside of the liner in a uniform up-
ward gradient to a high point where the geomembrane is
penetrated. Here the gases are vented, flared, or cap-
tured for energy use.
While not widely implemented, geotextiles with adequate
planar transmissivity could serve this purpose. The
design-by-function concept is again used to select the
proper material, where FS = qaiiow/qreqd. The required
gas flow rate is very site specific, but is available in the
landfill gas literature. The allowable gas flow rate uses an
adapted form of ASTM D-4716, but with radial rather than
parallel flow (see Figure 3-8).
REFERENCES
1. U.S. EPA 1988. U.S. EPA guide to technical resour-
ces for the design of land disposal facilities. EPA
guidance document: Final covers on hazardous
waste landfills and surface impoundments. EPA/530-
SW-88-047.
2. Koerner, R.M., ed. 1989. Durability and aging of
geosynthetics. Elsevier Applied Science Publishers.
332 pp.
3. Koerner, R.M., J.A. Bove, and J.P. Martin. 1984.
Water and air transmissivity of geomembranes, Vol.
1. No. 1, pp. 57-74.
33
-------
Load
Thickness
gages
Fabric-
Water
head
4.0
c 3.0
E
a 2.0
I
1.0
= 3.5 psi
U, =0.1 psi
0 500 1000 1500 2000 2500
Stress (lb./ft?)
Figure 3-8. Allowable gas flow as adapted from ASTM D-4716 (3).
34
-------
CHAPTER 4
DURABILITY AND AGING OF GEOMEMBRANES
POLYMERS AND FOUNDATIONS
There is an almost infinite variety of polymeric
geomembranes or flexible membrane liners—that can be
used in landfill cover situations. The major groups are:
• Thermoset elastomers (which are rarely used due to
seaming difficulties)
• Thermoplastic
• polyvinyl chloride (PVC)
• chlorosulphonated polyethylene (either nonrein-
forced or reinforced)—CSPE or CSPE-R
• ethylene interpolymer alloy (reinforced)—EIA-R
• very low density polyethylene (VLDPE)
• high density polyethylene (HOPE)
• Bituminous (which are rarely used in the United
States)
Geomembranes in the thermoplastic group are currently
most frequently utilized. These particular polymer resins,
however, are used in a formulation to arrive at the final
compound. Table 4-1 presents the formulations typically
used.
Table 4-1. Typical Formulations of Geomembranes
Geomembrane Resin
Type (%)
Carbon Black
and/or
Plasticizer Filler Additive*
PVC
CSPE
EIA
VLDPE
HOPE
45-50
45-50
70-80
96-98
97-98
35-40
2-5
10-25
0
0
10-15
45-50
5-10
2-3
2-2.5
3-5
2-4
2-5
1-2
< 1
'refers to antioxidant, processing aids, and lubricants.
When assessing durability and aging of membranes,
each component of the compound within a particular
geomembrane must be addressed. Obviously, those for-
mulations that contain no plasticizers or fillers have a
less complicated set of mechanisms to consider, e. g. ,
VLDPE or HOPE. There is a wealth of information in the
literature on "resin" behavior, but little on the durability of
specific geomembrane formulations. For this reason, this
section will treat each possible degradation mechanism
individually, before dealing with synergistic effects and
lifetime prediction methods.
MECHANISMS OF DEGRADATION
Eight separate mechanisms of degradation are described
in this section. Some of the major ones, such as
ultraviolet degradation, can be eliminated by soil covering
and others, such as radiation or chemical degradation,
are not very likely due to the geomembrane's position in
the closure system above the waste.
Ultraviolet Degradation
By virtue of its short wavelength components, sunlight
can enter into a polymer system and (with sufficient ener-
gy) cause chain scission and bond breaking. Figure 4-1
shows the wavelength spectrum of visible and ultraviolet
radiation. Superimposed on the figure are the most sensi-
tive wavelengths of some commercially used polymers
for geosynthetic materials:
• PE<300nm
• PET<325nm
• PP<370nm
The mechanism of ultraviolet degradation is well under-
stood and two approaches are taken to minimize its ef-
fects. Carbon black is added to the formulation, as a
blocking or screening agent, and chemical stabilizers are
added as scavenging agents. To eliminate degradation of
geomembranes in closure situations, however, the
geomembrane should be placed, seamed, and inspected,
and then covered with soil soon afterward.
Radiation Degradation
Clearly, radiation degradation is only of concern if there is
radioactive waste in the facility. Both y-rays and neutrons
within waste can degrade polymers and cause chain
scission and bond breaking. While there is a real concern
for high-level and transuranic wastes, low-level waste is
substantially less radioactive, and may not be a problem.
Chemical Degradation
Various chemicals can be aggressive to certain types of
geomembranes. For this reason, EPA has developed an
entire test protocol called EPA 9090 testing for assessing
35
-------
270 290 310 330 350 370 390 410 430 450 470 490 510 530 550 570 590
|f?|pET fpp Wtv«l»nglh
Figure 4-1. Wavelength spectrum of visible and ultraviolet radiation.
chemical resistance. While testing is necessary for liners
beneath the waste in contact with leachate, the closure
liner should only interface with water, which comes from
seepage through the cover soil placed above it. Thus,
chemical degradation generally should not be an issue
with the thermoplastic geomembranes used in landfill
closures.
Swelling Degradation
All polymers swell when exposed to liquid, including
water. Generally, HOPE swells the least, PVC swells the
most, and the other geomembranes fall between these
two extremes. The swelling process is largely reversible
and does not necessarily lead to degradation. However,
swelling may cause secondary actions that could lead to
other synergistic effects.
Extraction Degradation
If one, or more, of the components of a geomembrane
formulation are extracted, the remaining material will ob-
viously be compromised. For example, if swelling leads
to bond breaking of the plasticizer within a PVC formula-
tion, the plasticizer could be extracted over time. This
phenomenon would decrease the elongation capability of
the geomembrane with respect to tension, tear, and
puncture modes of failure. The closest tests available to
estimate extraction are the following:
• ASTM D3083 for water extraction.
• ASTM D1203 for volatile loss.
Other tests, or test procedures, could be required on the
basis of a site-specific and material-specific basis.
Delamination Degradation
For geomembranes that are fabricated in layers, e. g. ,
scrim reinforced or multi-ply types, there is a possibility
that liquid can enter between the layers causing
delamination and premature failure.
To prevent delamination, the edges of multi-layered
geomembranes should be properly sealed in the factory
and have no scrim exposed. If the sheets are trimmed in
the field, the exposed edges should be "flood coated"
with a heavy bodied solvent. An ASTM test on ply ad-
hesion, ASTM D413, should be performed on the
material.
Oxidation Degradation
Oxidation of polymers caused by the gases or liquids in-
terfacing with the geomembrane is unavoidable. The
oxygen, over time, will enter into the polymer structure
and can react with various components in the particular
formulation. All geomembranes (and all geosynthetics)
are subject to this type of oxidation mechanism. The fol-
lowing equation illustrates this mechanism for
polyethylene degradation:
R • + O2 -> ROO '
ROO ' + RH-> ROOM + R '
where R' = free radical
ROO ' = hydroperoxy free radical
RH = polymer chain
ROOM = oxidized polymer chain
The rate of the reaction is very site and polymer specific
and is usually addressed experimentally. The testing pro-
cedure will be discussed in the section on accelerated
testing methods.
To minimize the oxidation reaction, the polymeric for-
mulation contains various anti-oxidants which scavage
(i.e. , neutralize) the free radicals. The amount of anti-
oxidants that can be added, however, is limited, and
once it is utilized, the oxidation process will proceed
depending on site-specific and geomembrane-specific
conditions.
Biological Degradation
Microorganism degradation of the "resin" portion of the
various geomembrane formulations is probably not a
problem. There is no literature available concerning bac-
36
-------
terial or fungal attack of high molecular weight resins.
Microorganisms may, however, interact with the plas-
ticizers and/or fillers used in certain geomembranes. Two
ASTM tests can be used to detect this type of degrada-
tion: G21 deals with resistance of plastics to fungi and
G22 is the complementary test for bacterial resistance.
Higher forms of biological life, like burrowing animals,
may be a much more serious problem. A muskrat, or
other small mammal, interested in burrowing through a
geomembrane, could easily do so. The hardness of the
geomembrane versus the animal's teeth structure, force,
and hardness need to be considered. If such animals are
in the vicinity of the landfill, one might consider using a
rock "bio-barrier" above the geomembrane as per EPA
guidance.
SYNERGISTIC EFFECTS
The eight degradation mechanisms discussed in the pre-
vious section—ultraviolet, radiation, chemical, swelling,
extraction, delamination, oxidation, and biological—can
also interact to cause numerous complex effects. In addi-
tion, there are three situations that should be addressed
in any discussion on aging of polymeric materials:
elevated temperature, applied stresses, and long ex-
posure.
Elevated Temperature
All of the previously mentioned degradation processes
will be more severe at higher temperatures than at lower
temperatures. Activation energy, as will be discussed in
the section on Arrhenius modeling, is clearly a function of
elevated temperature. Thus, a given formulation of a
specific geomembrane will have a shorter lifetime in the
southern states (all other things than temperature being
equal) than in the northern states. A quantification of this
amount, however, requires experimentation and ap-
propriate modeling of the situation.
Applied Stresses
It seems intuitive that the lifetime of a geomembrane in a
relaxed state would be different than that of the same
geomembrane under stress. Thus, modeling of a given
situation should somehow take stresses into account. At
a minimum, compressive stresses should be assessed;
however, tensile and shear stresses (for liners on side
slopes) and out-of-plane bending stresses (for liners in
closures over subsiding waste) are also likely and should
be considered. Simulating the magnitude and the type of
stresses is very difficult and requires making many as-
sumptions.
Long Exposure
Landfill closures are designed (at the minimum) for 30-
year postclosure care periods. Beyond this time frame,
the facility's status is questionable and inquiries often
arise as to ultimate ownership and responsibility for main-
tenance and repairs. To alleviate some of these con-
cerns, it would be useful to quantify in the planning
stages how long a geomembrane will last before its
properties are degraded beyond serviceability. The
longer the geomembrane's lifetime, the easier it will be to
deal with these issues.
ACCELERATED TESTING METHODS
A number of accelerated testing methods in the open
literature predict the lifetime of polymeric materials. Many
of these sources are available in the plastic pipeline
literature from the Gas Research Institute, the Plastic
Pipe Institute, and related organizations. The reference
list for "Long-Term Durability and Aging of
Geomembranes" (see Appendix B) also contains many
useful sources of information.
Stress Limit Testing
Stress limit testing to obtain plastic pipe "design stress" is
fairly well developed and widely implemented. It requires
simulated environmental testing with sections of pres-
surized capped pipe at constant temperature. Figure 4-2
shows typical results where the service time of the pipe is
selected and the design stress is obtained from the ex-
perimental curve.
Rate Process Method for Pipe
In this experimental method for polyethylene pipe, con-
stant stress tests are conducted at different elevated
temperatures. The design life is selected, intersected with
the site-specific temperature response curve, decreased
by a suitable factor-of-safety, and then extended to ob-
tain the allowable stress. The graph in Figure 4-3 shows
pipe behavior at 80°C, 60°C, and 20°C. The behavior at
20°C must be extrapolated as per the details in the paper
by Koerner, Halse, and Lord in Appendix B. (See Appen-
dix B, "Long-Term Durability and Aging of
Geomembranes.")
Rate Process Method for Geomembranes
The rate process method (RPM) can be used to test
geomembranes if the site-specific conditions are properly
modeled. The curve in Figure 4-4 gives notched constant
load test data in an aggressive incubation liquid. The
method is not developed for utilization at this time.
Arrhenius Modeling
Arrhenius modeling is the method most widely used by
chemists and polymer engineers to predict the lifetime of
polymeric materials. This type of modeling assumes that
elevated temperature can be used to simulate time at a
site-specific (and lower) temperature. This assumption is
sometimes referred to as a temperature-time superposi-
tion concept. Figure 4-5 shows an example of a possible
test device. In Arrhenius modeling, measuring a suitable
reaction rate of the polymer at different temperatures
produces a linear plot on an inverse temperature graph.
37
-------
STRESS LIMIT TESTING
100 i
(B
£: 10 1
0}
Q.
O
o
I
Design Stress
1 10 100 1000 10000 100000 1000000
Failure Time (hrs.)
Figure 4-2. Stress limit testing for plastic pipe.
RATE PROCESS METHOD FOR GEOPIPES
log a (Mpa)
Callow = 6.5
20°C
60°C Ductile
Ductile
Figure 4-3. Rate process method for testing pipe.
The slope of the curve (if it is linear) can be used to ex-
trapolate to the site-specific temperature. The calcula-
tions then follow along the lines given below.
Information regarding polyethylene shielding of electric
cables is used here to demonstrate this technique. The
reaction rate illustrated in Figure 4-6 is for 50 percent
strength reduction of the original and unaged value using
an impact test. (Any one of a number of tests could have
been used.) The slope of the Arrhenius plot shown is:
In10"5-ln10"2
Eact
R
0.00247-0.00198
= -14,000 CK)
To predict low-temperature behavior, consider the
elevated temperature data point at:
1/T = 0.00213
log time
50 years
T
T
469°K
196°C
and project this reaction rate to the site-specific tempera-
ture of 90°C. Using the Arrhenius equation,
Rn = Eact [ 1 _ 1
Rr2 R Ti T2
where Rn
Rr2
Eaci/R =
Ti
T2
reaction rate at temperature Ti
reaction rate at temperature T2
slope of experimental curve
experimental (test) temperature
site-specific (desired) temperature
38
-------
Hydraulic Load Device
Air (7)
Liquid Level
Sight Glass
"- — 'Record
^4
^-*
1
o
o
o
o
0
o
0
er a
Leachate t=-
Recirculation =
Pump ^
S-^*-^f-^*s
LiQUI
^-— *
T.
1
>^>-
d
^
J
»
1
^
-------
.00001
u.0018
0.0020 0.0022 0.0024
1 / TEMPERATURE (1 /°K)
0.0026
Figure 4-6. Reaction rate for impact testing of polyethylene shielding.
(Figure 4-7c) curves. By superposition of the proper
temperature response curve and the appropriate strain
response curve (laboratory and field), one can possibly
project the lifetime of the considered geomembrane
under three possible assumptions:
• No additional stress relaxation, curve (a)
• Intermediate stress relaxation, curve (b)
• Full stress relaxation, curve (c)
These results are each shown on the graph in Figure 4-
7d. This technique is potentially useful, but requires a
relatively large amount of experimental and field data.
SUMMARY AND CONCLUSIONS
Durability and aging of geomembranes (and all geosyn-
thetics) are important issues, especially when consider-
ing the situation of landfill covers beyond the 30-year
postclosure care period. Fortunately, most degradation
processes are eliminated or greatly reduced by burying
the geomembrane in soil soon after installation. Also, be-
cause the interfacing liquid is water and not leachate, as
with the liner beneath the waste, there is little problem
with chemical degradation. Long-term oxidation,
however, is a degradation mechanism that can only be
retarded (via anti-oxidants), but not eliminated, and, thus,
is a focal parameter for experimental modeling.
Of the predictive models that have been reviewed, the
Arrhenius modeling technique, which is under active in-
vestigation, is in the most widespread use. Equally inter-
esting is the multi-parameter approach, but this method is
much less developed. Whatever techniques are used,
they are only laboratory prediction methods. Field feed-
back is necessary to establish better insight into degrada-
tion and aging issues involving polymeric geomembranes
and other related geosynthetic materials.
40
-------
100
90
• 0
70
O 5°
100
9
CONSTANT STRESS TESTS
(e.g., NCLT or SCLT)
CTILE REGION
i.i I L L.
10 1 10 10 10
TIME (HOURS)
~~5 T~
10 10 10
1. I
0.1 3
0.7 W
Q.
0.8 5
03
0.25
..2
0.15 5
CONSTANT STRAIN TESTS
(I.e., STRESS RELAXATION)
100 1000 10000
YEARS
10 i to 10" 10
TIME (HOURS)
100 1000 10000
YEARS
COUPLE LAB AND FIELD
STRAIN GAUGE DATA
1 to 10 10
TIME (HOURS)
100 1000 10000
YEARS
90
80
70
60
S 50
_l
UJ
LL
O 40
UJ
30
20
SUPERPOSITION OF SITE SPECIFIC
• CONSTANT STRESS CURVE,
• CONSTANT STRAIN CURVE, and
• FIELD STRAIN CURVE
(a) no addl relax.
(b) intermed. relax
(c) full relax.
ir* it
•1
0.9
0.8
0.7
0.6
0.5
0.4
»
Q.
LL
o
0.25 J
i
02 g
0.15 ^
1 10 10 10
TIME (HOURS)
10 10
10 10
1 0
10
100 1000 10000
YEARS
Figure 4-7. Experimental and field-measured response curves for multi-parameter lifetime prediction.
41
-------
CHAPTER 5
ALTERNATIVE COVER DESIGNS
INTRODUCTION
The hazardous waste landfill cover designs developed
and published by the U.S. Environmental Protection
Agency (EPA) are generic in nature and intended to meet
the regulatory criteria for covers on a national basis. This
section reviews these designs to determine alternatives
that would be acceptable to the regulatory community.
This discussion will also review designs being considered
by other agencies, such as the Nuclear Regulatory Com-
mission (NRC).
SUBTITLE C
The basic acceptable generic design for hazardous waste
landfills incorporates natural soil (clay), geomembrane,
drainage, and vegetation layers. This generic design,
however, does not take into account site-specific con-
cerns, such as siting in arid areas where rainfall is very
low. Under these conditions, a barrier layer composed of
both a natural soil (clay) and a geomembrane layer
probably would not be effective. The natural soil layer is
designed to be placed "wet-of-optimum" to achieve the
minimum hydraulic conductivity. When placed in a rela-
tively dry environment, this layer will dry and crack,
making it less effective. In selected cases, the newer
bentonite blankets may be an acceptable alternative.
From a technical standpoint, the geomembrane may be
the only barrier necessary. Site-specific considerations
such as settlement/subsidence, environmental exposure,
and other physical conditions may influence the thickness
of the geomembrane required.
In selected cases, a vegetation layer alone may be
demonstrated (via the Hydrologic Evaluation of Landfill
Performance (HELP) model—see Chapter 8) to meet the
criteria. In this case, a thicker soil layer may be required
to assist in establishing the natural vegetation and to act
as a storage reservoir for the infrequent but high intensity
rainfall.
In summary, the design criteria were established for a na-
tional generic design. EPA is always interested in review-
ing alternative designs that are innovative and utilize
site-specific information. These alternative designs
should be demonstrated to be equivalent in performance
to the generic design proposed by EPA.
SUBTITLE D
While EPA has proposed some generic design con-
siderations, Subtitle D facility designs will most likely be
approved by individual states. Cover designs should be
incorporated into the overall facility design, taking the
bottom liner and liquids management strategy into ac-
count. Depending on site-specific considerations, designs
based on natural soils as well as designs that resemble
multilayer Subtitle C designs will be developed.
Municipal solid waste landfills usually require a daily
cover of natural soils or other alternative materials. One
possible use for postconsumer paper or unsaleable glass
or glass culls may be for daily cover. Some enterprising
individuals have developed a product made from
shredded paper mixed with other proprietary ingredients
that can be blown onto the surface of the waste to meet
the requirements of daily cover. Foams and other
materials have also been developed and evaluated for
performance. Each of these materials has to be
evaluated economically for site-specific use but may
have advantages technically. For example, if the liquids
management strategy for a landfill includes leachate
recirculation, blown-on materials may allow more
homogeneous distribution of the leachate. Another ad-
vantage is that blown-on materials will lose their barrier
qualities as soon as the next layer is placed in the facility.
Natural soils, on the other hand, do tend to act as bar-
riers, which may cause leachate to seep out the side of
the final cover.
Whatever alternate materials are used, they should be
demonstrated to meet the technical requirements for
daily cover.
CERCLA
CERCLA or Superfund cover designs are more complex
from the standpoint of jurisdiction, where ARARs (dis-
cussed in Chapter 1) play an important part in selecting
the final design. A multilayer cover system may be most
environmentally desirable; however, other site-specific
considerations may allow other types of designs. For ex-
ample, early CERCLA covers have been constructed by
regrading existing cover material, and adding small
amounts of cover soil (usually about 6 inches) and, in
43
-------
some cases, a rock armor. Due to the inequality and with
the ARARs ruling, compliance with a RCRA multilayer
has been more acceptable. Site-specific design changes
have been approved after they were demonstrated to
meet the intent of the regulations.
OTHER COVER DESIGNS
The Department of Energy (DOE) and the NRC are both
considering cover designs for landfills containing low
level radioactive wastes. In general, these designs are
comparable to the EPA's multilayer design, with some
notable exceptions. One of the main criteria differences is
based on the fact that DOE and NRC designs have to
last for thousands of years due to the type of waste they
are covering. The long-term nature of their designs has
minimized the use of geosynthetics, since geosynthetics
are thought to have a finite service life.
DOE will soon publish results of a study designed to
develop an all natural soil cover system with a long ser-
vice life (1). The study considered what type of soil would
best qualify for each design aspect. A matrix was
developed from which completed matrix designs will be
proposed.
NRC also has been reviewing conceptual designs that
use natural soils and have long life. Three cover designs
are currently under investigation: (a) resistive layer bar-
rier, (b) conductive layer barrier, and (c) bioengineering
barrier (2). These designs are being assessed in large
(21 x 14 x 3 m [70 x 45 x 10 in.] each) lysimeters in
Beltsville, Maryland. The resistive layer barrier, shown in
Figure 5-1, consists of compacted natural soils or clay.
The resistive layer depends on the low hydraulic conduc-
tivity of the compacted layer to minimize any potential
moisture interaction with the waste.
The conductive layer barrier, shown in Figure 5-2, makes
use of the capillary barrier phenomena to increase the
moisture content above the interface and to divert water
away from and around the waste (3). The capillary barrier
is established when coarse grain soils are sandwiched
between fine grain sediments. Experiments have shown
that the greater the contrast in the permeability between
the two layers, the more effective the barrier. A second
fine grain soil layer would direct water away from the
gravel layer under saturated conditions.
It should be noted that NRC considers these two concep-
tual designs unacceptable where appreciable subsidence
may take place (2). This failure potential in the above two
designs necessitated the development of an easily
reparable surface barrier to be used until major settle-
ment/subsidence activities had ceased. The surface bar-
rier could be easily repaired during the
settlement/subsidence time period, after which a more
permanent barrier could be installed. The bioengineering
management cover system (Figure 5-3) was the result.
This cover system utilizes a combination of engineered
enhanced runoff and stress vegetation, e.g., Pfitzer
junipers, growing in an overdraft condition to control deep
water percolation through cover systems. Stress vegeta-
tion are grasses, trees, and shrubs that can survive when
under stress, such as lack of water. Early results from the
field indicate this system to be very effective in controlling
liquid movement into or out of the waste management
unit.
Figure 5-1. Resistive layer barrier.
44
-------
Figure 5-2. Conductive layer barrier.
Juniper (Pfttzer)
12 to 14'on Center
V
Access Tube (Neutron Probe)
2" Aluminum-
Gutter
-Gutter
Steel Drums (55 gal.,
1/3 Filled with Gravel)-7
3/4'Washed Gravely , ,
Native
SoU
- 4-20 mil. Vinyl Liners
Between 5 Layers of Geotextile
~t
- Instrumont
Pit
T
Vinyl
Geotextile
Figure 5-3. Side view of bioengineered lysimeter. Surface runoff is collected from both engineered surface and soil sur-
face. Soil moisture content is measured with neutron probe. Water table is measured in well.
REFERENCES
1. Identification and ranking of soils for UMTRA and
LLW disposal facility covers. U.S. Army Corps of En-
gineers, Waterways Experiment Station. Un-
published.
2. O'Donnell, E., R.W. Ridky, and R.K Schulz. 1990.
Control of water infiltration into near surface LLW dis-
3.
posal units. Progress report on field experiments at a
humid region site, Beltsville, MD. Waste Manage-
ment '90 Tucson, Arizona, February.
Zunker, J.F. 1930. Das Verhalten des Bodens Zum
Wasser In: E. Blanck, ed. Handbuch Der Bodenlehr.
V. 6. Berlin: Verlag von Julius Springer, pp. 66-220.
45
-------
CHAPTER 6
CONSTRUCTION QUALITY ASSURANCE FOR SOILS
INTRODUCTION
Construction quality assurance (CQA) is critical for
producing engineered cover systems that will perform
satisfactorily. The critical CQA issues for soils used in
cover systems are:
• Control of the soil materials used to build various com-
ponents of the cover system.
• Control of subgrade preparation.
• Control of the placement and compaction of soils.
• Protection of the soil during and after construction.
• Use of test pads in the CQA process.
This chapter examines each of these CQA issues. An
EPA guidance document (1) provides general information
concerning CQA, including responsibilities of the contrac-
tor and inspector. In general, the purpose of CQA is to
provide observations and tests that assist in evaluating
whether the construction has been performed in accord-
ance with specifications. Accordingly, each CQA program
must be tailored to the specific construction specifica-
tions for a given project. The sections that follow discuss
general principles that should be considered when
developing a CQA plan.
MATERIALS
The most important and useful quality control (QC) tests
for soil materials used in cover systems are Atterberg
limits, percentage of fines, percentage of gravel, and the
maximum size of the largest stones or clods of clayey
soil.
Atterberg Limits
The liquid and plastic limits or Atterberg limits of a soil,
which are measured with ASTM Method D4318, can be
useful indicators of the suitability of a soil for a specified
purpose. The plasticity index (PI) of a soil is defined as
the liquid limit minus the plastic limit and is a measure of
the breadth of water content over which the soil behaves
plastically. For hydraulic barrier materials, the soil must
have adequate plasticity; otherwise, the material will be
too deficient in clay to serve as an adequate hydraulic
barrier. For drainage materials, the soil must be free of
clay—drainage materials typically have little or no plas-
ticity. By measuring the plasticity of a soil, Atterberg limits
provide a rapid and convenient means for assessing its
suitability for its intended purpose.
The Atterberg limits of a soil are measured in the
laboratory. Samples for testing can be taken from the
borrow area or from the final construction area. Ex-
perienced field engineers and technicians can often tell
just from examining and handling the soil whether it has
the appropriate Atterberg limits. With questionable soils,
or soils that are variable in the borrow area, it is helpful to
sample the borrow soils on a close grid of test pads and
wait for the test results before proceeding with further soil
processing or placement.
Percentage of Fines
The percentage of fine-grained material in a soil is
defined as the percentage on a dry-weight basis of soil
that will pass through the openings of a No. 200 sieve,
which are 0.075 mm (0.003 in.) wide. Material retained
on the No. 200 sieve is defined as the coarse-grained
fraction and material that will pass through the openings
is the fine-grained fraction. The percentage of fines may
be measured with ASTM Method D422 or D1140, with
sample preparation performed with Method D2217, if
necessary.
As with Atterberg limits, an experienced engineer or tech-
nician can often tell by visual observation whether the soil
has an adequate amount of fines (barrier materials). In
such cases, the QC tests serve primarily to build a formal
record of test results that verifies the observations made
by the field personnel. With sandy drainage materials, it
is usually difficult to determine by visual observation
alone whether the material has an excessive amount of
fines.
Percentage of Gravel
The soil for barrier materials cannot contain an exces-
sively large percentage of gravel. The gravel fraction is
determined by sieving the soil through a No. 4 sieve,
which has 4.76-mm (0.19-in.) wide openings. Sieving of
soil to determine the percentage of gravel is performed
with ASTM Method D422, a method similar to the test for
percentage fines. All material that will not pass through
the openings is defined as gravel, according to the
Unified Soil Classification System (ASTM Standard Prac-
47
-------
tice D2487). (Note: Sieve analysis via ASTM Method
D422 is one of several tests used for soil classification
via the procedure for interpretation of test data given in
Standard Practice D2487. Also, particles larger than 75
mm [3 in.] are cobbles.)
Maximum Size of Panicles or Clods
For barrier materials, the maximum size of stones in the
clayey soil cannot be too large. However, it is impractical
for field personnel to sieve large, representative samples
of soil to determine the largest particle size. In the field,
the problem is probably an occasional oversized stone,
which no formal sampling procedure is likely to detect.
Rather, observations by CQA personnel provide the best
opportunity to detect excessively large stones.
On occasion, the maximum size of soil clods may be
specified in the construction specifications. Again, sieving
wet, clayey soils to determine the clod size distribution is
impractical. Direct measurement of representative clods
by field personnel is probably the simplest and best way
to verify that the clods are not too large.
Requirements for Field Personnel
On large and important projects, where CQA is con-
sidered crucial to the overall success, a full-time inspec-
tion of soil excavation in the borrow area and continuous
classification of excavated soils are recommended. Soils
are variable materials, and the borrow area offers the
best opportunity to detect the presence of unsuitable
materials.
Frequency of Testing
Table 6-1 summarizes the frequency of testing recom-
mended by Daniel (2). These recommendations are
based primarily upon practices reported by Gordon,
Huebner, and Kmet (3) and reiterated by Goldman et al.
(4) for clay liners. Although the recommendations are in-
tended for low hydraulic conductivity liners, they are use-
ful for other soil materials, as well. Experience has shown
that even more frequent testing is helpful during the initial
phases of construction, because this is the period when
problems are most likely to occur.
CONTROL OF SUBGRADE PREPARATION
The subgrade must be properly prepared and compacted
before any component of the cover system that requires
significant compaction can be placed. Typically, the low
hydraulic conductivity soil barrier requires compaction
with heavy equipment, whereas the compaction of other
layers is much less important. Thus, CQA for subgrade
preparation is critical for the low hydraulic conductivity
component of a cover system but may not be necessary
for other components.
Table 6-2 summarizes the recommended tests for sub-
grade preparation. For low hydraulic conductivity soils,
the surface of a previously compacted layer of soil should
be disked ("scarified") prior to placing a new layer of soil.
Table 6-1. Recommended Materials Tests for Barrier
Layers (2)
Parameter Test Method Minimum Testing
Frequency
Percent Fines (1) ASTM D1140 1 per 1,000yd3 (2)(3)
Percent Gravel (4) ASTM D422 1 per 1,000 yd3 (2)(3)
Liquid & Plastic ASTM D4318 1 per 1,000 yd3 (2)(3)
Limits
Water Content ASTM 04643(5) 1 per 200 yd3 (2)(6)
Water Content (7) ASTMD2216 1 per 1,000 yd3 (7)(3)
Construction
Oversight
Observation
Continuous in borrow
pit on major projects;
continuous in placement
area on smaller projects
Notes:
1. Percent fines is defined as percent passing the No. 200 sieve.
2. In addition, at least one test should be performed each day
that soil is excavated or placed, and additional tests should
be performed on any suspect material observed by QA
personnel.
3. 1,000 yd3 = 836m3.
4. Percent gravel is defined as percent retained on the No. 4
sieve.
5. This is a microwave oven drying method. Other methods
may be used, if more appropriate. Any method used besides
direct drying via ASTM D2216 should be calibrated against
ASTM D2216 for the onsite solids.
6. 200 yd3 = 167m3.
7. Microwave oven drying and other rapid measurement
methods may involve systematic errors. Conventional oven
drying (ASTM D2216) is recomrrmeded on every fifth sample
taken for rapid measurement. The intent is to document any
systematic error in rapid water content measurement.
Requirements for scarification should be set forth in the
construction specification. The scarification should be ob-
served to ensure that the newly placed layer blends in
with the previously compacted layer.
SOIL PLACEMENT
Soils are placed in layers called lifts. The thickness of a
lift is typically specified in the construction documents.
For materials having low hydraulic conductivity, the thick-
ness of a lift should not exceed the specified value.
Otherwise, the lower portion of the lift may be inade-
quately compacted, the bonding of lifts is likely to be
poor, and the hydraulic conductivity could be larger than
desired. Control of lift thickness is critical for low-
hydraulic-conductivity, compacted soil liners. The loose
lift thickness can be checked visually near the edge of a
lift. The exact thickness of a loose lift can be measured
by digging a hole in the soil.
48
-------
Table 6-2. Recommended Tests and Observations on
Subgrade Preparation (2)
Parameter
Test Method
Minimum Testing
Frequency
Percent
Compaction (1)
ASTM D2922 or
ASTMDl556or
ASTM D2937 or
ASTMD2167
1 per acre (2)
Compaction Curve ASTM D698 (3) 1 per 5 acres
Preparation of
Previously
Compacted Lift
Observation
Full coverage
Notes:
1. Percent compaction is defined as the dry density of the
compacted soil divided by the maximum dry density
measured in the laboratory with a specified method of
compaction. The test methods listed are for measurement
of the dry density of the compacted soil.
2. In addition, at least one test should be performed each day
the construction personnel prepare subgrade by compaction.
3. Other laboratory compaction methodologies are often
employed.
4. 1 acre = 0.4 ha.
Fill elevations are usually controlled with grade stakes or
lasers; laser equipment is not currently in widespread
use. If grade stakes are used, care must be taken to
remove them and repair the resulting holes. The CQA in-
spector should make sure that grade stakes are not
buried in the cover system. To accomplish this, an inven-
tory system in which all grade stakes are numbered and
accounted for each day is recommended. One advantage
of ferrous metal grade stakes is that if inadvertently
buried in the cover system, they can be found with a
metal detector. The holes left by grade stakes should be
packed with soil liner material or bentonite tamped into
the hole in layers with a rod.
SOIL COMPACTION
Drainage Layers
Nominal compaction of drainage layers is usually ade-
quate. Rarely is it necessary to control the degree of
compaction of drainage materials. One potential problem
to avoid is "bulking" of wet or damp sands; compaction in
lifts will overcome such problems.
Of greater importance than the degree of compaction is
protecting drainage materials from contamination by
fines. Over-compaction of the drainage materials can
grind up soil and increase the amount of fines. However,
the specifications should not permit use of nondurable
materials that are easily broken down. Field personnel
should observe the amount of fines before and after com-
paction. If there is any question about grinding of the soil
during compaction, the percentage of fines should be
measured after compaction to confirm that the compac-
tion process has not increased the percentage of fines.
Barrier Materials
Quality control of barrier materials usually focuses heavi-
ly on water content and dry unit weight. A typical con-
struction specification might require that the soil be
compacted over a specified range of water content (e.g.,
0 to 4 percent wet of optimum) to a minimum dry unit
weight (e.g., 95 percent of the maximum dry unit weight
from standard Proctor compaction).
The methodology for determining the appropriate com-
paction criteria for CQA has recently been reviewed by
Daniel and Benson (5). Figure 6-1 shows the form of a
typical compaction specification. The "Acceptable Zone"
is based upon a specified range of water content and a
minimum dry unit weight. The zero air voids curve repre-
sents a theoretical upper limit above which points cannot
exist. Figure 6-1 represents the usual format for specify-
ing the compaction requirements for a barrier layer.
However, as indicated by Daniel and Benson, the usual
format represents historical practice for structural fills and
is not necessarily appropriate for low hydraulic conduc-
tivity soil liner or cover systems. The next several graphs
illustrate the problem with the usual form of specification.
Figure 6-2 contains a compaction curve and a plot of
hydraulic conductivity versus molding water content for
three compactive energies. The water content-dry unit
weight points are replotted in Figure 6-3 with solid sym-
bols used for those compacted specimens that had
hydraulic conductivities less than or equal to 1 x 10~7
cm/s and open symbols for specimens with a hydraulic
conductivity greater than 1 x 10~7 cm/s. The Acceptable
Zone, which encompasses the compacted specimens
with low hydraulic conductivity, has a much different
shape from the one shown in Figure 6-1. Figure 6-4
presents contours of values of water content and dry unit
weight that yielded certain hydraulic conductivities for
one particular soil. Also shown in Figure 6-4 is a modified
Proctor compaction curve and a typical specification that
might be written using the procedure suggested in Figure
6-1. In this case, a portion of the typical Acceptable Zone
contains soils with unacceptably large hydraulic conduc-
tivities. Use of the Acceptable Zone in Figure 6-4 based
on typical construction practice, i.e., the practice
sketched in Figure 6-1, does not ensure that the com-
pacted soils have low hydraulic conductivity.
The recommended procedure for defining a suitable
range of water content and dry unit weight is shown in
Figure 6-5. The procedure involves four steps:
1. The soil is compacted with three compactive ener-
gies that span the range of compactive effort ex-
pected in the field. The three energies recommended
by Daniel and Benson (5) are modified Proctor,
standard Proctor, and reduced Proctor. Reduced
49
-------
Zero Air Voids
Acceptable Zone
Specified
Range
w
opt
Molding Water Content, w
Figure 6-1. Traditional method for specification of acceptable water contents and dry unit weights (5).
Proctor is the same as standard Proctor except that
only 15 drops of the hammer (rather than the usual
25) are employed per lift. Approximately five to six
samples are compacted with each energy.
2. The specimens are permeated and the hydraulic con-
ductivity of each specimen is determined. Hopefully,
at least some of the test specimens will have
hydraulic conductivities that are less than the design
maximum value. Care should be taken to make sure
that the conditions of permeation appropriately simu-
late field conditions. For cover systems, it is par-
ticularly important that the confining stress used in
the laboratory tests is not significantly larger than the
value expected in the field (which is usually small for
cover systems).
3. The water content-dry unit weight points are
replotted, and an Acceptable Zone is drawn. Some
judgment may be necessary in drawing the Accept-
able Zone.
4. The final step is to modify the Acceptable Zone in
any appropriate manner to take into account other
variables besides hydraulic conductivity, e.g., sus-
ceptibility to desiccation damage, local construction
practices, or shear strength considerations. When the
Acceptable Zone is modified, it is only made smaller,
not larger. Figure 6-6 illustrates how one might com-
bine an Acceptable Zone based on hydraulic conduc-
tivity with one based upon shear strength to develop
a single, overall Acceptable Zone.
The lower limit of the Acceptable Zone will probably be
parallel to a line of constant degree of saturation or to the
line of optimums (Figure 6-7). (The line of optimums is a
curve that connects points of maximum dry unit weight
and optimum water content measured with different ener-
gies of compaction.) It may be possible to use a constant
degree of saturation or a line parallel to the line of op-
timums for the lower limit of the Acceptable Zone.
50
-------
CO
E
o
>s
^>
o
ID
c
o
O
o
CL
O)
10
10
10
10
10
10
-4
-5
-6
-7
-8
-9
H
\ Medium
u
High Effort
10
15
20
25
Molding Water Content (%)
120
110
•? 100
90
(B)
Low Compactive Effort
15
20
25
Molding Water Content (%)
Figure 6-2. Data from Mitchell et al. for silty clay compacted with impact compaction (6).
51
-------
120
o
Q.
D)
'CD
c
=)
110
100
90
Acceptable Zone
10 15 20 25
Molding Water Content (%)
Figure 6-3. Compaction data for silty clay (6); solid symbols represent specimens with hydraulic conductivity less than or
equal to 1x10"7 cm/sand open symbols represent specimens with hydraulic conductivity >1x10"7cm/s.
120
o
Q_
D)
"CD
110
100
90
Mod.
Proctor
Curve
10-5
Acceptable Zone Based on
Typical Current Practice:
TO > 0.9 ifci.max and
w = 0 - 4% Wet of w0pt
15
20
25
Molding Water Content (%)
Figure 6-4. Contours of constant hydraulic conductivity for silty clay compacted with kneading compaction (6).
52
-------
Molding Water Content
Molding Water Content
(B)
- Maximum Allowed k
X
-
Acceptable Zone
Modified to Account
lot Other Factors
Molding Water Content
Molding Water Content
(A) Determine compaction curves with three compactive efforts
(B) Determine hydraulic conductivity of compacted specimens
(C) Replot compaction curves using solid symbols for samples with adequately low hydraulic
conductivity and open symbols for samples with a hydraulic conductivity that is too large
(D) Modify Acceptable Zone based on other considerations such as shear strength or local
construction practices
Figure 6-5. Recommended procedure.
The water content of the soil at the time the soil is com-
pacted has a significant impact on almost all engineering
properties of the soil. For instance, compaction of the soil
at a low water content leads to a strong, low-compres-
sibility soil that is not as vulnerable to desiccation crack-
ing as wetter soils. However, dry soils are brittle and
crack easily, e.g., if there is differential settlement of the
underlying waste. Soils compacted at high water contents
are softer, more compressible, and more vulnerable to
damage from desiccation. Wet soils, however, are ductile
and can accommodate more differential settlement
without cracking than dry soils can. The water content
range specified for construction should reflect a careful
consideration of these, and possibly other, variables. It is
critical that the CQA program ensure that the soil is com-
pacted to the proper water content.
The procedure summarized in Figure 6-5 was applied to
a cover system constructed at the Oak Ridge Y-12
Operations as described by Daniel and Benson (5). The
compaction curves for one of the two types of soils
(called Type A soils) are shown in Figure 6-8. Hydraulic
conductivities are plotted in Figure 6-9. For this project,
the design called for compaction of the soil to achieve a
hydraulic conductivity of 1 x 10"7 cm/s or less. The Ac-
ceptable Zone, shown in Figure 6-10, shows the range of
water content and dry unit weight that met this require-
ment. In the field, technicians measured the water con-
tent and dry unit weight, checked to see if the point
plotted within the Acceptable Zone, and made a pass-fail
decision based on whether the point was within or out-
side of the Acceptable Zone. If the point was outside the
Acceptable Zone, either the soil was compacted more or
the soil was removed, reprocessed, and recompacted.
For soil bentonite mixes, the following procedure is
recommended:
1. Mix batches of soil at different bentonite contents,
e.g., 0, 2, 4, 6, 8, 10, and 15 percent bentonite (dry
weight basis).
2. Develop standard Proctor compaction curves for
each bentonite content.
3. Compact samples with standard Proctor procedures
at a water content 2 percent wet of optimum for each
bentonite content.
4. Permeate the soils prepared in Step 3 and develop a
plot of hydraulic conductivity versus bentonite con-
tent.
5. Decide how much bentonite to use based on data
from Step 4, taking into account construction
variability. Usually more bentonite is used than Step
4 indicates is necessary, because, in the field, the
bentonite will not always be mixed as uniformly with
the soil as it was in the laboratory.
6. For the average bentonite content expected in the
field, utilize the procedures described earlier and
summarized in Figure 6-5.
The procedures discussed in the preceding pages for
determining an appropriate range of water content and
dry unit weight involve laboratory tests. The compaction
53
-------
Table 6-3. Recommended Tests and Observations on
Compacted Soil for Barrier Layers (2)
Parameter
Test Method
Minimum Testing
Frequency
Water Content (1)
Water Content (3)
Density (4)
Density (Note 5)
Number of
Passes
Construction
Oversight
ASTMD3017or
ASTM D4643
ASTMD2216
ASTM D2922 or
ASTM D2937
D1556
Observation
Observation
5/acre/lift (2)
1/acre/lift (3)
5/acre/lift (2)
1/acre/lift (5)
1/acre/lift (2)
Continuous
Notes:
1. ASTM D3017 is a nuclear method and D4643 is microwave
oven drying. Direct water content determination (ASTM
D2216) is the standard against which nuclear, microwave,
or other methods of measurements are calibrated for onsite
soils.
2. In addition, at least one test should be performed each day
soil is compacted, and additional tests should be performed
in areas for which QA personnel have reason to suspect
inadequate compaction.
3. Every fifth sample tested with ASTM D3017 or D4643 also
should be tested by direct oven drying (ASTM D2216) to aid
in identifying any significant, systematic calibration errors
withD3017orD4643.
4. ASTM D2922 is a nuclear method and D2937 is a drive ring
method. These methods, if used, should be calibrated
against the sand cone (ASTM D1556). Alternatively, the
sand cone method can be used directly.
5. Every fifth sample tested with D2922 or D2937 also should
be tested (as close as possible to the same test location)
with the sand cone (ASTM D1556) to aid in identifying any
systematic calibration errors with D2922 or D2937. The
sand cone method may be used in lieu of D2922 and
D2937.
6. 1 acre = 0.4 ha.
procedures used in the laboratory should simulate field
compaction as closely as possible. There is always a
possibility, however, that field construction will produce
macro-scale features (e.g., poor bonding between lifts)
that cannot be duplicated with laboratory compaction.
Large-scale in situ hydraulic conductivity testing of field-
compacted soil is recommended for a test pad; the pur-
pose of such tests is to verify that hydraulic conductivity
objectives can be met on the field scale. Test pads are
discussed further later in this chapter.
Table 6-3 lists the tests and observations recommended
for ensuring that the soil is properly compacted. Gordon
et al. provides the basis for the recommended frequency
of testing (3). Periodic calibration checks are recom-
mended for tests that are performed with microwave
ovens, nuclear devices, drive rings, or other equipment
that may introduce a small, systematic bias in the test
results. Systematic measurement errors, especially for
water content, must be documented.
In Table 6-3, it is recommended that observations of the
number of passes of the compaction equipment be peri-
odically determined. The compactive energy delivered to
low hydraulic conductivity materials has a large influence
on the hydraulic conductivity of the materials after com-
paction. If too little compactive energy is delivered to the
soil, e.g., because too few passes of the compactor are
made over the soil, then the hydraulic conductivity may
not be as low as desired.
Some individuals place great emphasis on hydraulic con-
ductivity tests performed on "undisturbed" samples taken
from a compacted lift of low hydraulic conductivity
material. One sampling procedure is to push a thin-
walled sampling tube (sometimes called a "Shelby tube")
into the soil with a backhoe, as shown in Figure 6-11.
However, with this procedure, the sampling tube usually
rotates during the push (Figure 6-12), which leads to un-
acceptable disturbance of the soil sample. This sampling
procedure is strongly discouraged. A better procedure is
to use a thin-walled sampling tube that is only about 23
cm (9 in.) long. (The tube should never be pushed more
than about 23 cm [9 in.] into the soil because stiff, com-
pacted soils usually plug the sampling tube if a longer
push is used.) As shown in Figure 6-13, a hydraulic jack
is placed on top of a sampling tube. The jack is used to
push the sampling tube straight into the soil. A backhoe
can be used, but only as a reaction for the hydraulic jack
as shown in Figure 6-14. Sampling procedures described
in ASTM Practice D1587 should be followed. Procedures
for preserving and transporting samples should be in ac-
cord with ASTM Practice D4220.
There are several potential problems with laboratory
hydraulic conductivity tests performed on "undisturbed"
samples of the liner material for CQA purposes:
• If there are any rocks or stones in the soil, it may be
virtually impossible to obtain a representative sample
for testing. When stones are present, the sampling
tube drags the stones through part of the soil sample,
which damages the sample. Many samples may have
to be taken and discarded before a sample that does
not contain too many stones is obtained. However, the
value of testing a sample that contains almost no
stones, when most of the soil does contain stones, is
questionable.
• Small samples of soil may not be representative. If
there are cracks, zones of poor compaction, or other
hydraulic defects, the chances that an occasional
small sample will detect those defects are remote. Just
because laboratory hydraulic conductivities are low
54
-------
.c:
O
"CD
Overall Acceptable
Zone
Acceptable Zone
Based on Hydraulic
Conductivity
Acceptable Zone
Based on Shear
Strength
Molding Water Content
Figure 6-6. Use of hydraulic conductivity and shear strength data to define a single, overall acceptable zone (5).
does not necessarily mean that the field values are
also low.
• Laboratory hydraulic conductivity tests take from 1 day
to 1 week to complete. The value of this type of test for
CQ purposes is minimized by the long time required to
obtain results. If the completed lift is left exposed while
the CQA team awaits the results of hydraulic conduc-
tivity tests, the whole process may be counterproduc-
tive.
There is no widespread agreement about how CQA offi-
cials should deal with the problems noted above for
laboratory hydraulic conductivity tests. The problems are
noted for the reader's information, and it is hoped that
CQA planners will develop strategies for dealing with
these difficulties.
Finally, any holes made in the soil must be sealed.
Quality assurance personnel should visually inspect the
sealing of some of the holes made for QC testing.
PROTECTION OF A COMPLETED LIFT
Visual observations are recommended to determine if
adequate measures have been taken to protect each lift
of soil from desiccation, freezing, or other damaging for-
ces. Additional tests, e.g., water content tests, should be
required if there is any question that the soil may have
been damaged after compaction.
55
-------
o
Q.
-C
O
'CD
o
0.
D)
'
-------
O
Q.
0>
'(I)
-^
'c
Z)
130
120
110
100
90
\
\\
. Zero Air Voids
"Curve
D Red. Proctor
O Std. Proctor
A Mod. Proctor
10
15
20
25
30
Molding Water Content (%)
Figure 6-8. Compaction curves for Type A soil from East Borrow area at Oak Ridge Y-12 operations project (5).
East Borrow Area
Type A Soil
D Red. Proctor
O Std. Proctor
A Mod. Proctor
40
Figure 6-9. Hydraulic conductivity versus molding water content for Type A soil from East Borrow area at Oak Ridge Y-12
operations project (5).
25 30 35
Molding Water Content (%
57
-------
o
Q.
D)
"CD
Acceptable Zone
D Red. Proctor
O Std. Proctor
A Mod. Proctor
Molding Water Content (%)
Figure 6-10. Acceptable zone for Type A soil from East Borrow area at Oak Ridge Y-12 operations project (5).
SAMPLING PATTERN
The CQA plan should detail the procedures for selecting
locations where samples will be taken. A random pattern,
or a sampling pattern utilizing a grid system, is recom-
mended. However, the CQA personnel should have the
authority to request additional tests at any questionable
location.
TEST PADS
Occasionally, test pads are constructed for the purpose
of verifying that materials and methods of compaction will
achieve the desired results, e.g., low in-field hydraulic
conductivity. If a test pad yields adequate results, the
CQA team should utilize the test pad in the CQA
program. One of the purposes of QC testing and QA ob-
servations should be to ensure that the actual cover is
built to standards that equal or exceed those used in
building the test pad.
If a test pad is built, it may be desirable to have a proce-
dural construction specification. The test pad is used to
demonstrate that the construction procedure is ap-
propriate. If the construction procedure is specified, a
critical objective of the CQA program should be to make
an adequate number of observations to verify that the
specified procedure has been followed.
One or more in situ hydraulic conductivity tests are usual-
ly performed on the test pad. Testing procedures are dis-
cussed by Daniel (7) and Sai and Anderson (8). The
most widely used method of measurement is the sealed
double ring infiltrometer (SDRI). Testing procedures for
the SDRI are given in ASTM Method D5093. A bevy of
other tests (water content, dry unit weight, Atterberg
limits, etc.) is usually part of the testing protocol, as well,
to provide full documentation of the test pad.
OUTLIERS
Soils are variable materials. It is inevitable that the soil
materials will fail to meet specifications at some points in
the soil mass. If all QC tests pass, this does not mean
that the soil everywhere meets the project specifications;
it just means that enough tests were not performed to lo-
cate the occasional outlier.
Occasional outliers are not necessarily harmful. For bar-
rier layers, the soils are constructed in multiple lifts in part
as a result of recognition that soils are variable and that
the compaction process is not perfect. For drainage
layers, occasional pockets of low hydraulic conductivity
materials will not harm the performance of reasonably
thick drainage layers—permeating fluids will simply go
around the low hydraulic conductivity material.
58
-------
Figure 6-11. Pushing of thin-walled sampling tube with a backhoe.
.*»* ,'•
Figure 6-12. Tilting of sampling tube during push.
59
-------
Figure 6-13. Placement of hydraulic jack on top of sampling tube.
Figure 6-14. Use of backhoe as a reaction for hydraulic jack.
60
-------
At least two approaches may be utilized for dealing with
outliers:
• The usual procedure is not to allow any outliers.
However, for the reasons noted above, this approach
is not realistic and frequently causes some manipula-
tion of tests or test results to ensure that unrealistic
specifications are met. In addition, the CQA team at
the end of the project may be left trying to justify the in-
significance of the occasional outlier (after the con-
struction is complete).
• One possible solution is to permit an occasional outlier.
Usually, if a small percentage of tests fail, the effect of
the outliers is nil. Such a specification is more realistic
and tends to discourage manipulation of sampling
locations, tests, or test results to meet unrealistic
specifications.
If a test does fail and it is believed that the failure repre-
sents inadequate materials or inadequate construction
procedures, the extent of the failed area must be defined
and the area must be repaired. The specifications should
prescribe how the area to be repaired will be determined.
SUMMARY
Figure 6-15 is a checklist of critical parameters for low
hydraulic conductivity barrier materials. A checklist for
drainage materials is shown in Figure 6-16. The items
shown on these checklists are intended to ensure that the
materials of construction are adequate and that ap-
propriate methods of construction have been utilized. The
CQA process involves a combination of QC testing and
observation by qualified personnel.
REFERENCES
1. U.S. EPA. 1986. Technical guidance document: con-
struction quality assurance for hazardous waste land
disposal facilities. Office of Solid Waste and Emer-
gency Response, Washington, DC, EPA/530-SW-86-
031.
2. Daniel, D.E. 1990. Summary review of construction
quality control for compacted soil liners. In: R.
Bonaparte, ed. Proceedings, Waste Containment
Systems: Construction, Regulation, and Perfor-
mance. New York: American Society of Civil En-
gineers, pp. 175-189.
3. Gordon, M.E., P.M. Huebner, and P. Kmet. 1984. An
evaluation of the performance of four clay-lined
landfills in Wisconsin. In: Proceedings, Seventh An-
nual Madison Waste Conference on Municipal and In-
dustrial Waste, Madison, Wl. pp. 399-460.
4. U.S. EPA. 1988. Design, construction, and evaluation
of clay liners for waste management facilities.
Washington, DC. EPA/530/SW-86/007F.
5. Daniel, D.E. and C.H. Benson. 1990. Water content-
density criteria for compacted soil liners. Journal of
Geotechnical Engineering, Vol. 116, No. 12, pp.
1811-1830.
6. Mitchell, J.K., D.R. Hooper, and R.G. Campanella.
1965. Permeability of compacted clay. Journal of the
Soil Mechanics and Foundations Division ASCE, Vol.
91, No. 4, pp. 41-65.
7. Daniel, D.E. 1989. In situ hydraulic conductivity tests
for compacted clay. Journal of Geotechnical En-
gineering. Vol. 115, No. 9, 1205-1226.
8. Sai, J.O. and D.C. Anderson. 1990. Field hydraulic
conductivity tests for compacted soil liners. Geotech-
nical Testing Journal. Vol. 13, No. 3, pp. 215-225.
61
-------
CRITICAL VARIABLES FOR LOW HYDRAULIC CONDUCTIVITY LAYER
Material:
Minimum Liquid Limit =
Minimum Plasticity Index =
Maximum Particle Size =
Maximum Percentage of Gravel =
Minimum Percentage of Fines =
Water Content/Density Defined (Y/N)
Maximum Clod Size =
Lifts:
Scarify Surface Before Placing (Y/N)
Maximum Loose Lift Thickness =
Maximum Completed Lift Thickness =
Compactor:
Minimum Weight =
Type of Roller Drum
Compaction:
Minimum Number of Passes
Protection:
Protection from Dessication & Freezing (Y/N)
Figure 6-15. Checklist of critical variables for CQA of low hydraulic conductivity compacted soil used in a cover system.
62
-------
CRITICAL VARIABLES FOR DRAINAGE LAYER
Material:
Maximum Percentage of Fines
Maximum Particle Size =
Lifts:
Maximum Loose Lift Thickness =
Compactor:
Maximum Weight =
Type of Roller Drum
Compaction:
Desirable Number of Passes
Grinding of Soil;
Visual Inspection? (Y/N) _
Protection:
Protection from Contamination by Fines? (Y/N)
Figure 6-16. Checklist of critical variables for CQA of drainage materials used in a cover system.
63
-------
CHAPTER 7
CONSTRUCTION QUALITY CONTROL FOR GEOMEMBRANES
PRELIMINARY DETAILS
A number of preliminary steps must be taken to ensure
that optimal construction quality control (CQC) and con-
struction quality assurance (CQA) can be achieved at the
landfill site. These steps involve manufacture, fabrication,
storage at the factory, shipment, and storage at the site
of the geomembrane. This chapter will focus on
CQC/CQA of geomembranes, but all geosynthetics can
be viewed in a similar manner.
Manufacture
The first consideration in quality control is that the
geomembrane resin, and its entire formulation, must be
appropriate to the site where it will be installed. Ap-
propriateness can be evaluated using EPA 9090 chemi-
cal compatibility testing, or by direct comparison to a
local, state, or federal specification, or to a stand-
ardization group. The material may have to be chemically
"fingerprinted" in some situations to assure that the
delivered geomembrane has identical characteristics to
the approved test samples.
The thickness, width, and length of the geomembrane
also must be verified. This is best done at the
manufacturer's plant, since shipment costs are high and
receiving the wrong geomembrane at the job site can
result in uncomfortable arguments and inconvenient
delays.
Other details such as the diameter and strength of the
windup core, protective covering (if appropriate), and
proper marking and identification need to be assured.
Fabrication of Panels
For certain types of geomembranes, such as polyvinyl
chloride (PVC), chlorylsulfonated polyethylene-reinforced
(CSPE-R), and ethylene interpolymer alloy-reinforced
(EIA-R), the relatively narrow rolls (about 1.8-m [6-ft]
wide) are fabricated into wider panels of about four to six
roll widths. Panel fabrication requires factory seaming,
usually either dielectric or bodied solvent. There should
be a thin sheet of plastic over the seams to prevent one
layer from sticking to the next.
The completed panels are accordion folded in two direc-
tions, wrapped in a heavy cardboard box, and placed on
wooden pallets for shipment. Proper marking and iden-
tification at this stage is necessary to ensure proper
delivery.
Storage at Factory
Geomembranes stored at the manufacturer's or
fabricator's facility should be elevated off the ground.
They also should not be stacked so high as to deform the
lower rolls or layers; warm climates are particularly
troublesome in this regard. An enclosed storage space is
recommended.
Shipment
Rolls or pallets of geomembranes are usually shipped by
truck to sites in the contiguous 48 states. When shipped
in closed trailers, the geomembranes should be loaded
and unloaded by lifting rather than by pushing and pull-
ing. Front-end loaders equipped with long rods (called
"stingers") are used for rolled geomembranes and forklift
loaders are used for palletized geomembranes.
In cases where stacking of the geomembranes might be
of concern, the delivery trailer should be inspected at the
job site for squashed rolls or crushed boxes.
Storage at Site
Unless the geomembrane is used directly as it comes off
the shipping trailer, a safe storage area should be
provided. The rolls of geomembrane should be elevated
off the ground or at least placed on a dry soil area that
does not contain vegetation, stumps, or other sharp ob-
jects. Covering is usually not necessary providing the
geomembranes are installed within a short period of time.
Palletized geomembranes should also be stored on site
on dry, level ground with similar considerations.
When the geomembranes are to be stored on the site for
months or longer, they should be covered and/or have an
enclosure around them for protection.
SUBGRADE PREPARATION
The subgrade must be prepared according to the site-
specific plans and specifications. Thus, line and grade
must have been established and verified before any
geomembrane is brought into the facility and positioned.
There can be no sharp objects of any kind such as grade
stakes, tools, stones, or equipment beneath the
geomembrane.
65
-------
Ruts caused by the compaction equipment or by the
geomembrane placement equipment must be leveled by
hand. Ruts are particularly troublesome if they freeze in
their uneven profile. They must be leveled before the
geomembrane is placed by waiting until the ground thaws
or by breaking the uneven surfaces using pneumatic clay
spades and pavement breakers.
Geomembranes should never be placed in ponded water.
Such a procedure is indicative of a poor sequence of
construction operations. Seaming can never be ac-
complished under such conditions.
DEPLOYMENT OF THE GEOMEMBRANE
The geomembrane should be placed on the entire facility
in accordance with a predetermined roll or panel layout.
Layout is a site-specific consideration, but plans are
generally supplied by the geomembrane manufacturer,
fabricator, or installation firm. Usually the rolls or panels
are ordered in a particular direction.
The construction deployment equipment should be low
ground pressure units in comparison to the subgrade
stability. For landfill covers, this is sometimes difficult to
achieve because the waste beneath the construction
operations can be actively subsiding.
After a roll, or panel, is initially positioned or "spotted," it
usually must be shifted slightly for exact positioning. By
lifting the liner up and allowing air to get beneath some of
it, the liner can sometimes be "floated" into position. If
this is not possible (e.g., with thick geomembrane
sheets), the liner has to be shifted by dragging it along
the subgrade (or on the geosynthetic material beneath it).
The entire roll or panel must then be inspected for
blemishes, scratches, and imperfections. Finally, the roll
or panel is weighted down with sandbags to prevent
movement by wind or any other disturbance.
Complete rolls and panels have been captured by gusts
of wind and unceremoniously dumped in a corner of the
facility. The owner/operator should decide beforehand,
i.e., during preconstruction meetings, whether a liner in
this situation can be used again or not. A number of
damage scenarios should be described to anticipate such
problems, since they are not that uncommon.
GEOMEMBRANE FIELD SEAMS
There are many types of geomembrane seams, most of
which were developed for a particular type of
geomembrane. Table 7-1 shows methods of field seam-
ing currently in use.
Solvent Seams
Solvent seams use a liquid solvent placed by a squeeze
bottle between the two geomembrane sheets to be
joined, followed by pressure to make complete contact.
As with any of the solvent-seaming processes described
in this section, a portion of the two adjacent
geomembranes is actually dissolved, resulting in both li-
quid and gaseous phases. Too much solvent will weaken
the adjoining geomembrane, and too little solvent will
result in a weak seam. Therefore, great care is required
in providing the proper amount of solvent for the par-
ticular type and thickness of geomembrane. Care must
also be exercised in allowing the proper amount of time
to elapse before contacting the two surfaces, and in ap-
plying the proper pressure and duration of rolling. These
seams are used primarily on flexible thermoplastic
materials.
Bodied solvent seams are similar except that 8 to 12 per-
cent of the parent lining material is dissolved in the sol-
vent before the seam is made. The purpose being to
compensate for the lost material while the seam is in a li-
quid state and to create a viscous liquid that can be
brushed on the area to be bonded. Pressure is applied,
and heat guns or radiant heaters are used to aid the
process.
A solvent adhesive uses an adherent left after the solvent
dissipates. The adhesive thus becomes an additional ele-
ment in the system. Sufficient pressure must be used to
affect an adequate seam. Most thermoplastic materials
can be seamed in this manner.
Contact adhesives have a wide applicability to most
geomembrane types. The solution is applied to both
mating surfaces by brush or roller. After reaching the
proper degree of tackiness, the two sheets are placed on
top of one another, and pressure is applied by a roller.
The adhesive forms the bond and is an additional ele-
ment in the system.
Vulcanizing tapes and adhesives are used on very dense
thermoset materials such as butyl and EPDM. In this
process, an uncured tape or adhesive containing the
polymer base of the geomembrane and crosslinking
agents are placed between the two sheets. Upon applica-
tion of heat and pressure, crosslinking occurs, which
gives the necessary bond. Factory seams are made in
large vulcanizing presses or autoclaves, while field
seams require a portable machine to provide the neces-
sary heat and pressure. Since thermoset geomembranes
are rarely used in landfill covers, the technique is not par-
ticularly important for our purposes.
Thermal Seams
There are a number of thermal methods that can be used
on thermoplastic geomembrane materials. In all of them,
the opposing geomembrane surfaces are truly melted
into a liquid state. Temperature, time, and pressure all
play important roles: too much melting weakens the
geomembrane and too little melting results in a weak
seam. The same care as is necessary for solvent seams
must be taken with thermal seams.
Hot air seaming uses a machine consisting of a resis-
tance heater, a blower, and temperature controls to blow
66
-------
Table 7-1. Overview of Geomembrane Field Seams.
Method
Seam
Configuration
Typical Rate8 Comments
Solvent
200 ft./hr. Requires tack time
Requires hand rolling
Requires cure time
Bodied
Solvent
150 ft./hr. Requires tack time
Requires hand rolling
Requires cure time
Solvent
Adhesive
150 ft./hr. Requires tack time
Requires hand rolling
Requires cure time
Hot Air
50 ft./hr. Good to tack sheets together
Hand held and automated devices
Air temperature fluctuates greatly
No grinding necessary
Hot Wedge
300 ft./hr. Single and double tracks available
Built in nondestructive test
Cannot be used for close details
Highly automated machine
No grinding necessary
Controlled pressure for squeeze-out
Ultrasonic
300 ft./hr. New technique for geomembrancs
Sparse experience in the field
Capable of full automation
No grinding necessary
Fillet Extrusion
100 ft./hr. Upper and lower sheets must be ground
Upper sheet must be beveled
Height and location are hand-controlled
Can be rod or pellet fed
Extrudate must use same polymer
compound
Air heater can preheat sheet
Routinely used for difficult details
Flat Extrusion
50 ft./hr. Highly automated machine
Difficult for side slopes
Cannot be used for close details
Extrudate must use same polymer
compound
Air heater or hot wedge can preheat sheet
am = ft x .3048
67
-------
air between two sheets to actually melt the opposing sur-
faces. Usually, temperatures greater than 260°C (500°F)
are required. Immediately following the melting of the sur-
faces, pressure is applied by rollers. For some devices,
pressure application is automated by counter-rotating
knurled rollers.
In the hot wedge or hot knife method, an electrically
heated resistance element in the shape of a wedge is
passed between the two sheets to be sealed. As it melts
the opposing surfaces, roller pressure is applied. Most of
these seaming units are automated in terms of tempera-
ture, speed of travel, and amount of pressure applied. An
interesting variation of the technique is the dual-hot-
wedge method, which forms two parallel seams with an
unbonded space between them. This space is sub-
sequently pressurized with air and any lowering of pres-
sure signifies a leak in the seam. Lengths of hundreds of
feet can be field tested in this one step. The hot wedge or
hot knife method will be more fully described in the sec-
tion on nondestructive seam testing.
Dielectric bonding is used only for factory seams of
flexible thermoplastic geomembranes. In this method, an
alternating current, at a frequency of approximately 27
MHz, excites the polymer molecules, creating friction and
thereby generating heat. This heat melts the polymer,
and when followed by pressure, results in a seam. A
variation of this method, called ultrasonic seaming, has
recently been introduced for the manufacture of field
seams on polyethylene liners.
Ultrasonic bonding utilizes a generated wave form of 40
kHz, which produces a mechanical agitation of the op-
posing geomembrane surfaces. Following the melting
process, a set of knurled wheels is used to mix and apply
pressure to the material.
Electric welding is yet another new technique for
polyethylene seaming. In this technique, a stainless steel
wire is placed between overlapping geomembranes and
is energized with approximately 36 volts and 10 to 25
amps current. The hot wire radially melts the entire
region within about 60 seconds, thereby creating a bond.
It is later used as a nondestructive testing method with a
low current and a questioning wire outside of the seamed
region.
Extrusion Seams
Extrusion (or fusion) welding is used exclusively on
polyethylene geomembranes. It is directly parallel to
metallurgical welding in that a ribbon of molten polymer is
extruded between or against the two slightly buffed sur-
faces to be joined. The extrudate ribbon causes some of
the sheet material to be liquefied and the entire mass
then fuses together. One patented system has a mixer in
the molten zone that aids in homogenizing the extrudate
and the molten surfaces. The technique is called flat
welding when the extrudate is placed between the two
sheets to be joined and fillet welding when the extrudate
is placed over the leading edge of the seam to be
bonded.
DESTRUCTIVE SEAM TESTS
After a field-seaming crew has seamed a given amount
of material, it is important to evaluate their performance.
One procedure is to cut out a sample, send it to a
laboratory, and test it until failure in either shear or peel
modes (see Figure 7-1). Another option would be to test
it directly at the field site. But considering a
geomembrane sheet layout, where should the seam be
tested and in how many places? Because each seam
sample becomes a hole that must be appropriately
patched and then retested, the number of field-seam
samples is commonly reduced to a bare minimum. Then
only the method of seaming is assessed, not its con-
tinuity. The method includes installation type, tempera-
ture, dwell time (time during which seam is under
pressure), pressure, and other operational details affect-
ing seam quality. Samples will ordinarily be taken at the
start of the seaming operations in the morning and after
the midday break. Thereafter, sampling can be done on a
random or a periodic basis. Haxo (1) recommends a fre-
quency of six samples per km (6/3,300 ft) of seam on a
random basis, or one sample per 150 m (1/500 ft) of
seam on a uniform basis.
There is much current discussion on what constitutes an
acceptable seam. Nearly everyone agrees that the seam
test specimen must not fail within the seamed region it-
self; that is, a failure must be a sheet failure on either
side of the seamed region. This is called a "film-tear
bond" failure. Engineers are not in agreement, however,
as to the magnitude of the force required for failure. For
SHEAR TEST
I
PEEL TEST
Figure 7-1. Shear and peel test for geomembrane seams.
68
-------
seams tested in a shear mode, failure forces of 80 to 100
percent of the unseamed sheet strength are usually
specified. For seams tested in a peel mode, failure forces
of 50 to 80 percent of the unseamed sheet strength are
often specified. These percentages underscore the
severity of peel tests as compared to shear tests. For as-
sessing seam quality, the peel test is preferable.
NONDESTRUCTIVE SEAM TESTS
Although the method of seaming must be properly as-
sessed, the test tells nothing of the continuity and com-
pleteness of the entire seam. It does little good if one
section of a seam has 100 percent of the strength of the
parent material, if the section next to it is missed com-
pletely by the field-seaming crew. Thus, this section dis-
cusses only continuous methods of nondestructive
testing (NOT). In each of these methods, the goal is to
check 100 percent of all seams (see Table 7-2).
The air lance method projects a jet of air at approximately
350 kPa (50 Ib/in.2) pressure through an orifice of 5-mm
(3/16-in.) diameter. The jet is directed beneath the upper
edge of the overlapped seam to detect unbonded areas.
When such an area is located, the air passes through,
causing an inflation and fluttering in the localized area.
This method only works on relatively thin (less than 45
mils [1.1 mm]), flexible geomembranes, and only if the
defect is open at the front edge of the seam, where the
air jet is directed. It is strictly a contractor/installer's tool
to be used in a CQC manner.
In the mechanical point stress or "pick" test, the tester
places a dull tool (such as a blunt screwdriver) under the
top edge of a seam. With care, an individual can detect
an unbonded area, because it is easier to lift than a
properly bonded area. This rapid test depends complete-
ly on the care and sensitivity of the person performing it.
Only relatively thick, stiff, geomembranes are checked by
this method. Detectability is similar to that using the air
lance, but both methods are very operator dependent.
This test also is to be performed only by the installation
contractor and/or geomembrane manufacturer. Design or
inspection engineers should use one or more of the tech-
niques discussed below.
Electric sparking is an old technique used to detect pin-
holes in thermoplastic liners. In this method, a high-vol-
tage (15 to 30 kV) current detects any leakage to ground
(through an unbonded area) by producing sparking. The
method is not very sensitive to overlapped seams of the
type generally used in geomembranes and is used only
rarely for this purpose. Today, the technique has been
revived in a somewhat varied form. In the electric wire
method, a copper or stainless steel wire is placed
between the overlapped geomembrane region and is ac-
tually embedded into the completed seam. After seam-
ing, a charged probe of about 20,000 volts is connected
to one end of the wire and slowly moved over the length
of the seam. A seam defect between the probe and the
embedded wire produces an audible alarm from the unit.
The method is strongly advocated by some installation
Table 7-2. Overview of Nondestructive Seam Tests.
Primary User
General Comments
Nondestructive
Test Method
air lance
pick test
electric wire
dual seam
(positive
pressure)
vacuum chamber
(negative
pressure)
ultrasonic pulse
echo
ultrasonic
impedance
ultrasonic shadow
electric field
acoustic sensing
Contractor
yes
yes
yes
yes
yes
-
yes
yes
Design Engineer
Inspector
yes
yes
yes
yes
yes
yes
yes
yes
Third-Party
Inspector
-
-
yes
yes
yes
yes
yes
Cost of
equipment
$200
nil
$500
$200
$1000
$5000
$7000
$5000
$20,000
$1000
Speed of
tests
fast
fast
fast
fast
slow
moderate
moderate
moderate
slow
fast
Cost of
tests
nil
nil
nil
mod.
very high
high
high
high
high
nil
Type of
Result
yes-no
yes-no
yes-no
yes-no
yes-no
yes-no
qualitative
qualitative
yes-no
automatic
yes-no
Recording
Method
manual
manual
manual
manual
manual
automatic
automatic
automatic
manual and
manual
Operator
Dependency
very high
very high
high
low
high
moderate
unknown
moderate
low
moderate
69
-------
firms, but the literature gives conflicting opinions when
comparing this method to vacuum box testing (discussed
below).
The pressurized dual seam method was mentioned ear-
lier in connection with the double-wedge thermal seaming
method. The air channel that results between the double
seam is inflated using a hypodermic needle and pres-
surized to approximately 200 kPa (30 Ib/in.2) for a length
of 30 to 300 m (100 to 1,000 ft). If no drop on a pressure
gauge occurs, the seam is acceptable; if a drop occurs, a
number of actions can be taken:
• The distance can be systematically halved until the
leak is located.
• The section can be tested by some other leak detec-
tion method.
• A cap strip can be seamed over the entire edge.
The test is an excellent one for long, straight-seam runs.
It is generally performed by the installation contractor, but
often with the designer or inspector observing the proce-
dure and assessing the results.
Vacuum chambers (boxes) are the most common form of
nondestructive test currently used by design engineers
and CQA inspectors. In the vacuum chambers method, a
1-m (3-fl) long box with a transparent top is placed over
the seam and a vacuum of approximately 17 kPa (2.5
Ib/in.2) is applied. When a leak is detected, the soapy
solution originally placed over the seam bubbles, thereby
reducing the vacuum. The vacuum is reduced due to air
entering from beneath the liner and passing through the
unbonded zone. The test is slow to perform and it is often
difficult to make a vacuum-tight joint at the bottom of the
box where the box passes over the seam edges. Due to
upward deformations of the liner into the vacuum box,
only geomembrane thicknesses greater than 30 mils
(0.75 mm) should be tested in this manner. It would be
difficult to test 100 percent of the field seams by this
method, however, because of the large number of field
seams and the amount of time required. The test could
also not inspect around sumps, anchor trenches, and
patches with any degree of assurance. The method is
also essentially impossible to use on side slopes, since
the downward pressure required to make a good seal
cannot be obtained (it is usually done by standing on top
of the box).
A number of ultrasonic methods are available for seam
testing and evaluation. The ultrasonic pulse echo method
is basically a thickness measurement technique and is
only used with semicrystalline geomembranes. In this
method, a high-frequency pulse is sent into the upper
geomembrane and (in the case of a good seam) reflects
off of the bottom of the lower one. If, however, an un-
bonded area is present, the reflection will occur at the un-
bonded interface. The use of two transducers, a pulse
generator, and a CRT monitor are required. The
ultrasonic pulse echo test cannot be used for extrusion
fillet seams, because of their nonuniform thickness.
The ultrasonic impedance plane method works on the
principle of acoustic impedance. A continuous wave of
160 to 185 kHz is sent through the seamed liner, and a
characteristic dot pattern is displayed on a CRT screen.
The dot pattern is calibrated to signify a good seam. The
method has potential for all types of geomembranes but
still needs additional development work.
The ultrasonic shadow method uses two roller
transducers: one sends a signal into the upper
geomembrane and the other receives the signal from the
lower geomembrane on the other side of the seam. The
technique shows an energy transmitted on the display
monitor. HOPE seams with received signals greater than
50 percent full-scale height were all found to be accept-
able by subsequent testing with destructive methods.
Received signals less than 20 percent full-scale height in-
dicated that the seams were not acceptable. The 50 to 20
percent range had mixed results. This technique can be
used for all types of seams, even those in difficult loca-
tions, such as around manholes, sumps, and appur-
tenances. It is best suited to semicrystalline
geomembranes, such as HOPE, and will not work for
scrim-reinforced liners.
The electric field test utilizes a liquid-covered
geomembrane to contain an electric field. For this
method to work, the entire bottom of the lined facility
must be covered with liquid, usually water. The depth can
be nominal, approximately 15 cm (6 in.). Electric field
testing cannot be used where water does not cover the
geomembrane, as on side slopes. This method uses a
current source to inject current across the boundary of
the liner. When a current is applied between the source
and remote current return electrodes, current flows either
around the entire site (if no leak is present) or bypasses
the longer travel path through the leak itself (when one is
present). Potentials measured on the surface are af-
fected by the distributions and can be used to locate the
source of the leak. These potentials are measured by
"walking" a probe in the water. The operator walks on a
predetermined grid layout and marks where anomalies
exist. After the survey is completed, these anomalies can
be rechecked by other methods, such as the vacuum
box. Electric field testing is currently commercially avail-
able.
Acoustic sensing is used in conjunction with the dual
seam air pressure test. It is an effective method to locate
a leak if air pressure is not maintained.
PENETRATIONS, APPURTENANCES, AND
MISCELLANEOUS DETAILS
The various details of a geomembrane landfill closure are
very important in making the entire "system" function as
intended. Clearly, gas venting must be addressed in any
70
-------
closure over biodegrading solid waste landfills. Gas vent- always keep in mind that the waste will subside over time
ing pipes are usually constructed with prefabricated and typically in a very nonuniform and random manner.
"boots" around PVC or HOPE pipes. The geomembrane
is then seamed to the pipe according to the proper techni- REFERENCE
que. If metal pipes are used, a stainless steel clamp -, Haxo HEJ 1986 Qua|jty assurance of
usually makes the connection. geom'embranes used as liners for hazardous waste
Other appurtenances and details might need to be con- containment. Journal of Geotextiles and
sidered and details in the open literature should be Geomembranes. Vol. 3, No. 4, pp. 225-248.
evaluated accordingly. The landfill owner/operator must
71
-------
CHAPTERS
HYDROLOGIC EVALUATION OF LANDFILL PERFORMANCE (HELP) MODEL FOR DESIGN
AND EVALUATION OF LIQUIDS MANAGEMENT SYSTEMS
INTRODUCTION
Liquids management systems are critically important for
limiting leachate generation and migration. Cover sys-
tems control leachate generation by restricting infiltration
of precipitation into the waste layer. Leachate collection
and liner systems restrict migration of leachate from the
waste containment site by limiting leakage through liners
and promoting leachate collection. This chapter looks at
using the Hydrologic Evaluation of Landfill Performance
(HELP) model in the design and evaluation of these sys-
tems.
This chapter presents a brief overview of typical liquids
management systems, a detailed description of the HELP
model, and an example application of the HELP model
simulating a complete landfill system.
OVERVIEW
Landfills typically contain two liquids management sys-
tems. The cover is the principal liquids management sys-
tem for controlling leachate generation. Leachate, as
evaluated by the HELP model, is any rainfall or snowmelt
that combines with liquids in the waste and moves by
gravity forces to the bottom of a landfill. During its migra-
tion through the waste, the liquid takes on pollutants that
are characteristic of the waste. As such, the leachate
quantity and quality is site specific and waste specific.
The HELP model generates estimates of leachate quan-
tity given site-specific descriptions of climate, cover
design, and initial moisture content of the waste and soil
layers. The model does not predict leachate quality or
any contribution to the leachate quantity by subsurface
inflow of ground water. Good landfill design and site
selection would minimize any contribution from ground
water.
Covers
Figures 1-1 and 1-2 (see Chapter 1, pp. 1-3 and 1-7)
show typical cover designs recommended in the U.S.
EPA technical guidance document Final Covers on Haz-
ardous Waste Landfills and Surface Impoundments (1).
The designs are composed of three layers for liquids
management—vegetation/soil or cobbles/soil top layer,
drainage layer, and geomembrane liner/low hydraulic
conductivity soil layer (hydraulic barrier layer). The other
components in the design serve to support or maintain
the functions of these three layers.
The topsoil layer should be designed to promote runoff
from major storms, provide storage for evapotranspira-
tion, and protect the hydraulic barrier layer from frost
drainage layers
geomembrane
geomembrane anchors
, Ueparate anchor trench tor each geoiynthttlc)
permeability loll
wait* * geomembrane
Figure 8-1. Cover and liner edge configuration with example toe drain.
73
-------
Filler Medium
O Drainage Material O •*• Oittn Pioes -*»-O
Leachste
Collection and
Removal System
Being Proposed as the
Leak Detection Svslem
Low Permeability Soil
Native Soil Foundation
Lower Component
(compacted soil)
(Not toSeelel
Figure 8-2. Schematic of a single clay liner system for a landfill.
penetration and desiccation. Above all other considera-
tions, the topsoil, if vegetated, should have good mois-
ture retention properties. A clayey soil promotes runoff,
slows drainage, and provides storage for evapotranspira-
tion; however, it also can promote desiccation to greater
depths and retard vegetative growth.
Runoff promotes erosion and degradation of the cover
system. Vegetation or cobbles impede erosion and sup-
port the long-term integrity of the cover. The thickness of
the topsoil layer must be sufficient to retain adequate
moisture for maintaining vegetation and to ensure that
frost and desiccation will not penetrate to the hydraulic
barrier layer. Typically, a thickness of 60 to 90 cm (2 to 3
ft) is sufficient, but the actual requirement is site and
design specific.
The drainage layer should be designed to reduce
leakage through the hydraulic barrier layer. This is ac-
complished by draining the zone of saturation above the
barrier to a collection pipe or toe drain, as shown in
Figure 8-1 (2). This design lowers the hydraulic head
driving the leakage and decreases the quantity of water
available to leak through the barrier. All drainage layers
should be designed to have free drainage to a collection
system to maximize the drainage for a given system. Ab-
sence of a toe drain at the base of the side slope of the
cover system also can contribute to slope failures.
The drainage layer also reduces root and animal penetra-
tion of the hydraulic barrier layer and provides additional
depth and a capillary break to lessen desiccation and
frost penetration of the barrier. As shown in Figures 1-1
and 1-2, a filter, either soil or geotextile, must be placed
above the drainage layer to decrease migration of fines
and prevent the fines from clogging the drain. Drainage
layers are not necessary at all sites since some sites may
not have sufficient rainfall and infiltration to produce
standing head on the hydraulic barrier for long durations.
The hydraulic barrier layer or liner should be designed to
minimize the infiltration of water into the waste layer over
the long term. The liner systems shown in Figures 1-1
and 1-2 are both composite systems, but municipal waste
landfills have commonly used only a low hydraulic con-
ductivity soil liner. Use of a geomembrane in conjunction
with low hydraulic conductivity soil greatly improves the
effectiveness of the barrier. In conventional cover
designs, the barrier layer is the most important layer in
controlling leakage through the cover system. All other
layers tend to serve mainly as layers that support and
maintain the barrier. Generally, except in arid or semiarid
areas, the topsoil layer cannot promote sufficient runoff
and evapotranspiration to prevent leakage. Drainage
layers typically cannot drain all of the water passing
through the topsoil layer before it reaches the barrier.
Once water stands on the barrier, leakage occurs at a
rate controlled by the barrier.
Leachate Collection/Liner Systems
The second liquids management system used in a landfill
is the leachate collection/liner system. A typical single
liner system used for leachate collection is shown in
Figure 8-2 (2). It consists of a drainage layer overlain by
a filter, either soil or geosynthetic, and a hydraulic barrier
layer composed of hydraulic conductivity soil. The soil
liner is frequently overlain by a geosynthetic
geomembrane to greatly improve its performance. The
performance of these layers for liquids management is
the same as described above for cover systems except
that the layers are controlling leakage from the landfill in-
stead of infiltration into the waste layer and leachate
generation.
Figure 8-3 is a schematic of a typical double liner system
used both for leachate collection and leakage detection
(2). The top drainage layer and geomembrane is the
primary leachate collection system. The bottom drainage
74
-------
layer and composite liner (low hydraulic conductivity soil
overlain by a geomembrane) is the secondary leachate
collection system and leakage detection system.
HELP MODEL
Background
The HELP model was developed by the U.S. Army En-
gineer Waterways Experiment Station for the U.S. EPA
Office of Solid Waste (OSW) to provide technical support
for the Resource Conservation and Recovery Act
(RCRA) and Comprehensive Environmental Response,
Compensation, and Liability Act (CERCLA) programs.
Development of the model began in 1982 and Version 1
was released for public comment in June 1984 (3,4). The
program was a mainframe computer model that ran on
the National Computer Center's IBM system. In 1986, the
program was modified to run on IBM-compatible personal
computers. Additional capabilities and refinements were
included in Version 2 of the model released in 1988 (5,
6). The most current version is Version 2.05, which was
released in July 1989. Version 3 of the model is currently
in preparation for release in 1991.
The HELP model is a quasi-two-dimensional, gradually
varying, deterministic, computer-based water budget
model. It is termed quasi-two-dimensional because it
contains a one-dimensional vertical drainage model and
a one-dimensional lateral drainage model coupled at the
base of lateral drainage layers or the top of liners. The
program computes free vertical drainage down to the top
of a liner, at which point the liner restricts drainage and a
zone of saturation develops. The models for lateral
drainage and leakage or percolation through the liner
then use the height of saturated material above the liner
to compute simultaneously the rates of lateral drainage to
collection systems and vertical leakage through the liner,
respectively. The model is termed gradually varying be-
cause the simulation progresses through time using
analyses that are assumed steady for each time period.
Version 2 of the model uses a time period of 6 hours. The
model is deterministic and numerical as opposed to
stochastic and qualitative. Finally, the HELP model is a
computer-based water budget model; that is, the model
uses a computer to apportion the precipitation and initial
moisture content into estimates of the following water
budget components: surface runoff, evapotranspiration,
changes in snow storage, changes in moisture content,
lateral drainage collected in each drain system, and
leakage or percolation through each liner system. Figure
8-4 shows a schematic of the processes and systems
modeled by the program. Daily, monthly, annual, and
long-term average water budgets can be generated.
The HELP model is a tool developed specifically to aid
permit evaluators and landfill designers in the evaluation
and comparison of alternative landfill designs. The model
was built to evaluate whether alternative designs perform
as well as the minimum technical guidance systems over
a long period of time. Therefore, its primary utility is for
tasks involving comparison of alternative designs and
sensitivity of design parameters. A secondary goal of the
model's development was the accurate prediction of
water budget components. Nevertheless, additional test-
ing, verification studies, and refinements are ongoing to
improve its accuracy. In general, the accuracy and
precision of the model is limited by uncertainty and
variability in the properties of material existing in landfills.
As such, simulation results would be expected to be best
used to rate relative merits of designs rather than to ac-
curately predict the water budget components.
The HELP model is a valuable tool for design and permit
evaluation; however, it requires the user to exercise good
judgment. In particular, the user must have a good under-
standing of landfill design, vegetative systems, and the
model to obtain reliable results and correct conclusions.
The user must ensure the integrity of the design and the
data because the model does not evaluate the data.
Filter Medium
Top Liner Boiiom Composite
(geomembrane) Liner
Primary Leachate / _ , . ,. ,
- ,, . / Low Ptrmaabihiy Soil
CoJItfcdon and /
Removal System /
/
Secondar
CoUect
Remove
Being Proposed as
Leak Detection Sys
Leachaie Native Soil Foundation
on and
System
he
em
U w " 1
^r??^?]
Leacnate
Collection
System
Sump
Upper
Component
(geomembrane)
Lower Component
(compacted soil)
(Not to Scale)
Figure 8-3. Schematic of a double liner and leak detection system for a landfill.
75
-------
RAINFALL/SNOW
I! M I
INTERCEPTION .TRANSPIRATION
SNOW
EVAPORATION
SNOW
ACCUMULATION
RUNOFF
SNOW
MELT
VERTICAL PERCOLATION
INTERCEPTION
EVAPORATION
PLANT GROWTH
DEPTH OF
HEAD
BARRIER SOIL
PERCOLATION
Figure 8-4. Simulation processes in the HELP model.
An example of a cover system lacking design and data
integrity is a two-layer system with the following charac-
teristics. The top layer is 5 cm (2 in.) of topsoil vegetated
with a good stand of grass. The lower layer is 5 cm (2 in.)
of compacted clay having a saturated hydraulic conduc-
tivity of 10~8 cm/sec. With this description, the HELP
model would predict a very large quantity of runoff and a
small quantity of evapotranspiration since the topsoil
layer has very little storage capacity. In addition, very lit-
tle leakage through the clay liner would be predicted be-
cause the saturated hydraulic conductivity of the liner is
so low. The actual results would be expected to be quite
different. Construction of 5-cm (2-in.) layers is not prac-
ticable, especially, construction of a thin lift of clay com-
pacted uniformly to achieve an effective hydraulic
conductivity of 10~8 cm/sec. Besides the difficulties in
construction, the layers would lack integrity to maintain
the described properties. Both layers would quickly form
desiccation cracks, producing much larger hydraulic con-
ductivities. As such, the leakage and evapotranspiration
would be much greater than predicted and the runoff
would be much less.
The primary anticipated use of the HELP model—to per-
form quick comparisons of the long-term performance of
alternative designs, often with very little data—was con-
sidered throughout its development. As such, the follow-
ing approach was taken to select simulation methods and
data entry options. Process simulation methods had to be
well-accepted techniques described in the literature that
are computationally efficient, require minimum data that
are readily available, and account for all major design
and climate conditions. Conservative assumptions are
made when necessary because of uncertainty. The term
"conservative" implies that any resulting error would tend
to result in an overestimation of vertical drainage or
leakage through liners. Options are provided to permit
use of data from a default data base or from user entry.
Guidance and recommendations are given for poorly un-
derstood parameters. In addition, the program is interac-
tive and user friendly and runs on IBM-compatible
personal computers to facilitate widespread use.
Process Simulation Methods
The HELP model was adapted from the Hydrologic
Simulation Model for Estimating Percolation at Solid
Waste Disposal Sites (HSSWDS) of the U.S. EPA (7, 8)
and the Chemical Runoff and Erosion from Agricultural
Management Systems (CREAMS) (9) and Simulator for
Water Resources in Rural Basins (SWRRB) models of
the U.S. Department of Agriculture (USDA) Agricultural
Research Service (ARS) (10). The following sections of
this chapter describe all of the principal hydrologic and
physical processes modeled by the HELP model, includ-
ing a discussion of the assumptions and limitations of the
models of the principal processes. Many of the processes
are shown on a schematic of a closed landfill profile in
Figure 8-4. Understanding of the processes, simulation
methods, and their assumptions and limitations is critical
for the proper application of the HELP model.
Infiltration
Daily infiltration into the landfill is determined indirectly
from a surface water balance. Infiltration equals the sum
of rainfall and snowmelt, minus the sum of runoff and sur-
face evaporation. Runoff and surface evaporation are in
part a function of interception. Precipitation on days
having a mean temperature below 0°C (32°F) is treated
76
-------
as snowfall and is added to the surface snow storage.
Decreases in snow storage occur by snowmelt and sur-
face evaporation.
Daily precipitation is an input parameter. Precipitation
data may be synthetically generated, specified by the
user, or selected from the default data base of historical
rainfall data. The synthetic weather generator will be
described later in this chapter.
Snowmelt is computed using a slightly modified version
of the simple degree-day method with 0°C (32°F) as the
base temperature (11). The in./degree-day snowmelt
constant was increased from 0.06 to 0.10, a value more
typical of open areas. In addition, the modification per-
mits a small quantity of snowmelt to occur at mean daily
temperatures between -5° and 0°C (23° and 32°F) to ac-
count for the variation in temperature during a day and for
the fact that landfills often have higher soil temperatures
because of heat generated from biodegradation. Snow-
melt contributes to runoff, evaporation, and infiltration.
Interception is modeled after the work of Morton (12). In-
terception approaches a maximum value exponentially as
the rainfall increases to about 0.5 cm (0.2 in.). The maxi-
mum interception is a function of the quantity of
aboveground biomass or leaf area index and is limited to
a maximum of 0.13 cm (0.05 in.). The interception
evaporates from the surface and decreases the evapora-
tive demand placed on the plants and soil column.
The HELP model uses the Soil Conservation Service
(SCS) curve number method for estimating surface
runoff, as presented in the Hydrology Section of the Na-
tional Engineering Handbook (11). The SCS curve num-
ber method is an empirical method developed for small
watersheds (about 12.1 to 202.4 hectares [30 to 500
acres]) with mild slopes (about 3 to 7 percent). The
method correlates daily runoff with daily rainfall for water-
sheds with a variety of soils, types of vegetation, land
management practices, and antecedent moisture condi-
tions (levels of prior rainfall). As applied, the technique
accounts for changes in runoff as a function of soil type,
soil moisture, and vegetative conditions. Version 3 of the
model will include a procedure to adjust the curve num-
ber as a function of surface slope since surface slopes
greater than 20 percent can produce significantly greater
runoff.
Many assumptions and limitations exist in the applica-
tion of this method in the HELP model, including the
following:
• The SCS curve number method is applicable for
landfills that are much smaller in area than water-
sheds. Verification studies have shown good agree-
ment between the predicted and observed cumulative
annual volume of runoff.
• Cumulative volume of runoff is independent of rainfall
duration and intensity since over a long simulation
period a variety of precipitation events will occur. The
predicted value represents an average of the
measured runoff for the typical variety of rainfall events
of a given quantity.
• No surface run-on from surrounding areas is permitted
by the model.
• Estimates of runoff greater than predicted by the SCS
curve number method are produced when the surface
soils are saturated or limit infiltration due to very low
hydraulic conductivity.
Evapotranspi ration
Evapotranspiration consists of three components:
evaporation of water from the surface, from the soil, and
from the plants. Each component is computed separate-
ly. Evaporation of water from the surface is limited to the
smaller of the potential evapotranspiration and the sum of
the snow storage and interception. The HELP model
uses a modified Penman method to compute potential
evapotranspiration. This method, developed by Ritchie
(13), is also used in the CREAMS program (9). The
potential evapotranspiration is a function of ground cover,
daily temperature, and daily solar radiation. Evaporation
of surface water decreases the evaporative demand
placed on the plants and soil column.
The HELP model uses Ritchie's method of evaporation
from soil (13) as applied in the CREAMS (9) and SWRRB
(10) models. The method uses a two-stage, square root
of time routine. In stage one, the soil evaporation equals
the evaporative demand placed on the soil column.
Demand is based on energy and is equal to the potential
evapotranspiration discounted for surface evaporation
and shading from ground cover. A vegetative growth
model is used to compute the total quantity of vegetation,
both active and dormant, which provides shading. In
stage two, evaporation from the soil column is limited by
low soil moisture and low rates of water vapor transport
to the surface by soil suction. Stage two soil evaporation
is a function of the square root of the length of time that
the soil has been in this dry condition.
The HELP model estimates plant transpiration in the
manner of the CREAMS and SWRRB models (9, 10)
whereby the potential plant transpiration is a linear func-
tion of the potential evapotranspiration and the active leaf
area index. The leaf area index of actively transpiring
plants is computed using a vegetative growth model that
accounts for seasonal variation in active and dormant
aboveground biomass and leaf area index. This model
was extracted from the SWRRB model developed by the
USDA ARS (10). See "Vegetative Growth" later in this
chapter for a complete description of this model.
Many assumptions and limitations exist in the application
of this three-component evapotranspiration method in the
HELP model, including the following:
77
-------
• The potential evapotranspiration is a function solely of
the energy available at the surface and, therefore, is
not a function of energy produced in the landfill, soil
temperature, wind, and humidity. As such, the program
uses a vapor pressure gradient that is a function solely
of mean daily ambient air temperature.
• A constant value is used for the albedo (fraction of in-
cident solar radiation that is reflected). The value is
typical for brown soils and grasses and is modified
only when the surface is covered with snow.
• The program uses a constant evaporative zone depth.
This depth is the maximum depth to which soil suction
can draw water to the surface. The depth is a function
of soil properties, design, vegetation, and climatic con-
ditions.
• Ritchie's two-stage soil evaporation method is ap-
plicable for all materials, not just soils.
• Synthetically generated daily temperature and solar
radiation values are sufficient for estimating potential
evapotranspiration.
• The vegetative growth model produces representative
leaf area indices and biomass estimates that are suffi-
cient to estimate interception, surface shading, and
plant transpiration.
Subsurface Water Routing
Subsurface water routing processes modeled by the
HELP model include vertical unsaturated drainage, per-
colation through saturated soil liners, leakage through
geomembranes, and lateral saturated drainage. In model-
ing these processes, the soil moisture of each layer of
the landfill profile being modeled is computed by sequen-
tial analysis proceeding forward through time. The soil
moisture controls the rate of subsurface water movement
by each of the subsurface processes, but the rates of
water movement by these subsurface processes yield the
resulting soil moisture. Consequently, the soil moisture
and rates of subsurface water routing are computed
simultaneously in the HELP model by an iterative
process after accounting for extractions by soil evapora-
tion and plant transpiration.
The HELP model simulates unsaturated vertical drainage
using a unit hydraulic pressure gradient approach
(saturated Darcy's law) where drainage occurs at a rate
equal to the unsaturated hydraulic conductivity. Under
this approach, vertical water routing is only downward ex-
cept in the evaporative zone where water is removed up-
ward by evapotranspiration. The unsaturated hydraulic
conductivity is computed by the Campbell equation using
Brooks-Corey soil parameters to define the shape of this
power function (14, 15). This approach incorporates the
moisture retention properties (capillarity) of the soil in the
determination. The model considers limited interactions
between layers of materials. As such, the model does not
allow drainage from one layer at a rate greater than the
maximum infiltration rate of the layer below it, allowing
placement of a lower hydraulic conductivity nonliner layer
below a layer of higher hydraulic conductivity.
Future versions may consider soil matrix interactions in
the recommendation of values for the evaporative zone
depth and in the selection of the moisture content where
the drainage from one layer into another will cease.
These additions will better model the physical situations
where fine-grained materials overlie coarse-grained
materials. In these situations, the coarse-grained material
may restrict the depth of evapotranspiration and the fine-
grained material may retain a higher water content before
draining into the coarse-grained material despite very low
water content in the coarse-grained material. Conversely,
coarse-grained material overlying fine-grained material
will restrict the transport of water vapor up from below for
evaporation but will freely drain to very low moisture con-
tents. These phenomena occur because the soil suction
of fine-grained material is much greater than that of
coarse-grained material.
Vertical drainage through saturated soil liners is termed
percolation in the HELP model. The barrier soil liners are
assumed to remain permanently saturated, but percola-
tion occurs only when there is a zone of saturation direct-
ly above the liner. Percolation is computed by Darcy's
law using the saturated hydraulic conductivity of the liner
material. The head loss gradient is equal to the average
head above the base of the liner divided by the thickness
of the liner.
Leakage through geomembranes is modeled as a reduc-
tion of the cross-sectional area of flow through the sub-
soil below the geomembrane. The rate of flow through
the leaking subsoil is computed as the percolation rate
through a saturated barrier soil liner. This method
provides good results for composite liners but is not very
good for just a geomembrane. Therefore, Version 3 will
include an improved leakage model for geomembranes
based on the work of Brown (16) and Giroud et al. (17,
18,19).
The HELP model simulates lateral drainage using a
steady-state analytical approximation of the numerical
solution of the Boussinesq equation (Darcy's law for
saturated lateral flow through unconfined porous media
coupled with the continuity equation). The analytical ap-
proximation was developed by converting the Boussinesq
equation into a nondimensional form and solving it for
two analytical solutions at the extremes in nondimen-
sional average saturated depth. These two solutions
were then fitted with the same value and slope in an ap-
proximation that covers the rest of the range of non-
dimensional depths. The approximation matched the
numerical steady-state solution of the Boussinesq equa-
tion within 1 percent of the predicted drainage rate. The
solution is not linear; therefore, the HELP model uses a
Newton-Raphson method to converge onto the solution
78
-------
of the nonlinear approximation. The model uses the
average depth of saturation in the approximation since
the HELP model is quasi-two-dimensional and, therefore,
cannot determine the saturated depth profile. The lateral
drainage, percolation, and leakage through the
geomembrane are solved simultaneously with the
average depth of saturation using an implicit solution
technique.
Many assumptions and limitations exist in the application
of the subsurface water routing methods, including the
following:
• All flow is considered to follow Darcy's law. As such,
the only driving force for water routing is gravity, and
all movement has a downward component.
• All layers of materials are spatially homogeneous and
uniform. All properties of the layers and materials not
related to soil moisture are assumed to remain con-
stant throughout the simulation.
• No subsurface inflow occurs. The layers being
modeled are above the surrounding water table or cut
off from it.
• Brooks-Corey relationships and the Campbell equation
are applicable for estimating the unsaturated hydraulic
conductivity of all types of materials.
• Percolation and leakage through liner systems occur
only when a zone of saturation (head) lies on the top
surface of the liner. The zone covers the entire area of
the liner and, therefore, percolation and leakage are
spatially uniform.
• The liner and drainage layers cover the entire area of
the landfill since the water routing in the vertical and
lateral directions is performed separately by one-
dimensional models.
• The liner system is permanently saturated and the
pressure head at the base of the liner is zero (liner is
above the water table).
• Synthetic liners (geomembranes) are impermeable ex-
cept through specific failure points (holes, punctures,
cracks, faulty seams, etc.) and function by reducing
the area of flow through the subsoil beneath the liners.
• The rate of leakage through geomembranes is mainly
a function of the number of holes, depth of saturation
above the liner, and the saturated hydraulic conduc-
tivity of the subsoil.
• The saturated depth profile (water table) in the lateral
drainage layer is typical of steady-state drainage and
gradually varies between different steady-state profiles
characteristic for different depths of saturation as the
simulation progresses.
• The lateral drainage rate can be reliably estimated
from the average depth of saturation throughout the
drain layer which is estimated from the average soil
moisture content of the drain layer.
• The depth of saturation at the edge of the drain layer
or at the collector is zero. Therefore, lateral drainage is
not retarded by standing water in the drain trench.
Vegetative Growth
The HELP model accounts for seasonal variation in ac-
tive and dormant aboveground biomass and leaf area
index through a general vegetative growth model. This
model was extracted from the SWRRB model developed
by the USDA ARS (10). The vegetative growth model
computes daily values of biomass and leaf area index
based on a maximum allowable value from input, daily
temperature and solar radiation data, and the beginning
and ending dates of the growing season. The maximum
value of leaf area index depends on type of vegetation,
soil fertility, climate, and management factors. The
program supplies typical values for selected covers;
these range from 0 for bare ground to 5.0 for an excellent
stand of grass. The HELP model maintains a data file
containing mean monthly temperatures and beginning
and ending dates of the growing season for 183 locations
in the United States. Vegetative growth is a linear func-
tion of the available solar radiation during the first 75 per-
cent of the growing season. Growth can be limited by
temperatures below 10°C (50°F) and low soil moisture.
Vegetative decay is modeled as exponential decay and is
also a function of temperature and soil moisture. The
decay process is modeled continuously, during both the
active growing and dormant seasons.
Accuracy
As stated previously, the primary purpose of the model is
to simulate alternatives for comparison, showing relative
value of alternatives and sensitivity of design parameters.
The secondary purpose is to quantify the water budget
components accurately. Generation of accurate predic-
tions requires good understanding of the model,
hydrologic processes, and landfill design and construc-
tion. In addition, accurate data describing the properties
and variability of all materials and the climate are essen-
tial.
Even with the best of data and knowledge of the model
and landfill, significant errors should be expected in the
estimates of the water budget components due to mini-
mum data requirements and limitations in the modeling
techniques. The following error bounds are believed to be
generally achievable when extensive and accurate data
are available to a knowledgeable user (ideal circumstan-
ces). The cumulative annual total for a water budget
component can typically be estimated within the larger of
the following error bounds: 25 percent of the total or 2
percent of the precipitation for the surface runoff com-
ponent, 10 percent of the total or 7 percent of the
precipitation for the evapotranspiration component, and
79
-------
10 percent of the total or 0.1 percent of the precipitation
for the percolation or leakage through liners component.
The error bound for cumulative annual lateral drainage to
collection systems is about 7 percent of the precipitation
and is equal to the sum of the other errors. Its error is de-
pendent on all other errors because those processes
occur first and any excess or shortfall in the extraction by
those processes controls the quantity of water available
for lateral drainage. These error bounds would be several
times larger when simulations are run with poor data and
a poor understanding of the model and landfill design.
Input Requirements
Climatological Data
Required Climatological data include daily precipitation,
daily mean temperature, daily solar radiation, maximum
leaf area index, growing season, and evaporative zone
depth. Daily precipitation data can be provided by three
options. The user may enter each value into the
precipitation data file. The second option is to select 5
years of daily precipitation data from a default data set
that has data for 102 cities throughout the United States.
The third option is to synthetically generate daily
precipitation data using a synthetic weather generator
available in the program. The program has statistical
coefficients describing the daily precipitation at 139 cities
throughout the United States. Improvement in the data
generation can be obtained by specifying the normal
mean monthly precipitation at the landfill site. The syn-
thetic weather generator was adapted from the WGEN
model developed by the USDA Agricultural Research
Service (20).
Daily mean temperature and daily solar radiation data are
synthetically generated by the program after the user
selects a location with similar weather from a set of cities
having available data. The program contains statistical
coefficients describing temperature and solar radiation
values at 183 cities throughout the United States. The
generation of daily temperature values can be improved
by specifying the normal mean monthly temperatures for
the landfill site. Similarly, the daily solar radiation values
can be improved by specifying the latitude of the landfill
site.
When executing the program, guidance is available from
a permanent data file for the remaining three climate-re-
lated parameters—growing season, maximum leaf area
index, and evaporative zone depth. Typically, the growing
season is that portion of the year when the mean daily
temperature is above about 11°C (53°F). The maximum
leaf area index is dependent on climate, soil fertility,
cover design, and management practices. Thick layers of
fine-grained soils have better fertility and moisture reten-
tion than thin layers and coarse-grained soils and, there-
fore, support better stands of vegetation. The evaporative
zone depth is dependent on climate, vegetation, and soil
properties. Representative evaporative zone depths for
silty, loamy topsoils are given in the program as a func-
tion of location and quantity of vegetation. Typical values
would be greater for thick clayey layers while the values
for sandy layers would be smaller. In addition, thin layers
of materials and the presence of synthetic material near
the surface also may restrict the evaporative zone depth.
Soil and Design Data
The second set of required data consists of a description
of each layer of material and a description of the landfill
design. Material descriptions can be selected from a
default set of material properties or specified individually
by the user. The material properties that must be
specified by the user are porosity, field capacity, wilting
point, and the saturated hydraulic conductivity. Porosity is
defined as the volume of voids in a layer of material (or
volume of water in a saturated layer) divided by the total
volume of the layer. Field capacity is defined as the
volume of water remaining in a layer of material after it
ceases to drain by gravity divided by the total volume of
the layer. It corresponds to the moisture content remain-
ing when the material exerts a soil suction of 1/3 atmos-
pheres. Wilting point is defined as the volume of water
remaining in a layer of material after a plant extracts as
much water as possible and goes into a permanent wilt,
divided by the total volume of the layer. It corresponds to
the moisture content remaining when the material ex-
hibits a soil suction of 15 atmospheres. User-specified
descriptions are recommended since material properties
vary greatly within a given soil classification.
The default set of material descriptions contains proper-
ties for 15 soil types ranging from coarse sand to high
plasticity clay. These 15 types also include fine sands,
loams, silts, and low plasticity clays. These soils also
may be specified as compacted, which causes the
porosity, field capacity, and saturated hydraulic conduc-
tivity to be lower. Compaction is recommended only for
the fine-grained materials. In addition to the 15 soils,
there are also two descriptions of very low hydraulic con-
ductivity soils suitable for liners and one description of
municipal waste with daily cover.
When using either option of describing materials, the
user has two options for initializing the moisture content
of the materials. The user may specify an initial moisture
content ranging from porosity to wilting point. This option
is used when the effects of changing moisture storage
are important in the water budget. Under the second op-
tion, the program initializes the moisture content of the
layer to the approximate long-term, steady-state value.
This option is used when changes in water storage are
unimportant or when the long-term, pseudo-steady-state
water budget is desired.
The landfill description consists of the SCS runoff curve
number, surface area, runoff area, and a description of
the layers including the number of layers, their order and
function, and their thickness. Four types of layers, based
80
-------
on how they function, are used by the model—vertical
percolation layers, lateral drainage layers, barrier soil
liners, and geomembranes with barrier soil. Vertical per-
colation layers are layers that serve no purpose other
than water storage. The topsoil layer and waste layers
are typical examples of this type. Lateral drainage layers
are layers designed to promote lateral drainage to a col-
lection system. They typically have very large saturated
hydraulic conductivities and are underlain by a sloped
liner. Barrier soil liners are layers of low hydraulic con-
ductivity porous material designed to restrict vertical per-
colation or leakage. A geomembrane with barrier soil is a
synthetic membrane underlain by subsoil and is used to
restrict vertical percolation and leakage. In addition, the
slope and drain spacing are needed for lateral drainage
layers and the liner leakage fraction is needed for
geomembranes.
Output Description
The output from the HELP model is a listing of the input
data followed by an account of the water budget com-
ponents in a tabular format. Information on precipitation,
surface runoff, evapotranspiration, lateral drainage from
each liner/drain system, percolation or leakage through
each liner and from the bottom of the profile, moisture
storage, snow accumulation, and depth of saturation on
the surface of liners is reported. A partial listing of the
output for an example application is presented in the next
section of this chapter. Simulation results are available at
several levels of detail. The cumulative quantity of the
water budget components and its variance are tabulated
on an average monthly and annual basis for the entire
simulation period and optionally on a daily, monthly, and
annual basis. In addition, peak daily results during the
entire simulation period and final moisture contents of the
layers are reported.
EXAMPLE APPLICATION
This section presents a simulation of the water balance
for the closed landfill illustrated in Figure 8-5. The landfill
has eight layers—a three-layer cover system, a waste
layer, and a four-layer double liner system. The cover
consists of a topsoil to support a fair stand of grass, a
sand layer to drain excess infiltration, and a clay liner.
The waste layer contains lifts of waste and daily cover.
The double liner system provides leachate collection and
leakage detection. It contains a sand layer for primary
leachate collection and a geomembrane for the primary
liner. A second sand layer serves as the subsoil for the
geomembrane and as the leakage detection or secon-
dary leachate collection layer. The lower liner is a com-
posite liner consisting of both a geomembrane and a clay
liner and is considered to be one layer. Many other
aspects of the design description required by the HELP
model are shown on the schematic.
The example profile can be simulated as having eight or
preferably nine layers. The layers are described in the
© LATERAL DRAINAGE LAYER qA.,n LATERAL DBAiNA
• CJMINU IFROM COVERI
(3) BARRIER SOIL LINER CLAY
(1) VERTICAL I
PERCOLATION LAYER TOPSOIL
VERTICAL
PERCOLATION
LAYER
WASTE
SAND
UjZ
xo
LATERAL DRAINAGE LAYER
LATERAL DRAINAGE
(LEACHATE COLLECTION)
FLEXIBLE MEMBRANE LINER
)
BARRIER SOIL LINER
Lei OPF"'^
MAXIMUM
CLAY OfU
PERCOLATION (LEAKAGE!
Figure 8-5. Typical hazardous waste landfill profile.
input from top to bottom; therefore, the first layer is the
topsoil layer and the last layer is the composite liner.
Eight layers are shown on the schematic but nine layers
were used in the simulation. The additional layer comes
by dividing the waste layer into two layers, the top of
which is thick and the lower of which is thin. Dividing thick
layers in this manner minimizes incontinuities in the solu-
tion. Incontinuities result from use of average moisture
contents with greatly different layer thicknesses and
occur when the zone of saturation above a liner extends
from a thin layer into a thick layer. The materials in the
profile were described using the default set of material
properties. A completed data form describing the landfill
materials and design is shown in Figure 8-6. The data
form is available in the user guide (5) and lists the data
requirements in the exact form and order that data entry
is made in the model.
Climatological input was entered using the default set of
rainfall data and the statistical coefficients and default
values for synthetic generation of daily temperature and
solar radiation values for Philadelphia, Pennsylvania.
Values for evaporative zone depth and maximum leaf
81
-------
DEFAULT SOIL AflD DESIGN DATA INPUT
Title:
/
Do you want the program to initialize the soil water?
Number of layers: V
Layer data:
Laver 1
(a) thickness c*y inches
(b) layer type {_ (1 or 2)
(c) liner leakage fraction (only for layer type 4) — (0 Co 1)
(d) soil texture number /£? (1 to 20)*
(e) compacted? (only for soil textures 1 to 15) A>Y} (Yes or No)
(f) initial soil water content (not asked if program is to initialize
the soil water or if layer type is 3 or 4) /^ oZ/V/* vol/vol
(must be between wilting point and porosity)
Laver 2
(a) thickness /3 inches
(b) layer type 2 (1 to 41
(c) liner leakage fraction (only for layer type 4) ~~ ~ (0 tc
(d) soil texture number /_ (1 to 20)*
(e) compacted? (only for soil textures 1 to 15) ////> (Yes or No)
(f) initial soil water content (not asked if program is to initialize
the soil water or if layer type is 3 or 4) /?,,?^yf^> vol/vol
(must be between wilting point and porosity)
Laver 3
(a) thickness ^p/4 inches
(b) layer type .7
(c) liner leakage fraction (only for layer type 4) • (0 to 1)
(d) soil texture number /£_ (1 to 20)*
(e) compacted? (only for soil textures 1 to 15) //?/• (Yes or No)
(f) initial soil water content (not asked if program is to initialize
the soil w.itpr or if layer type Is 3 or 4 ) /?. 'AiY?/7 vol/vol
(must be between wilting point and porosity)
Laver 4 Laver 5 Layer 6
(a) _ .fy.f (a) /J. (a)
(b) / (b) / (b)
(c) — (c) — (c)
(d) /? (d) //., (d)
(e) A^ (r) X& . (e)
(f) .^, ~-.PTJ (f)
Figure 8-6. Completed data form for landfill materials and design.
82
-------
(a)
(b)
(c)
(d)
(e)
(f)
(a)
(b)
(c)
(d)
(e)
(f)
If
Layer 7 Laver 8
<^ ( a ) /)
V- (b) J
/7. (?PfZ?S~ ( c )
/ (d) /
/_/S7 (e) /jff
/?, '/y '7 ( f ) /?, f9^7f
Laver 10 Laver 11
(a)
(b)
(c)
(d)
(e)
(f)
soil texture number of layer 1 is between
Type of vegetation: fT?j'/~
SCS runoff curve number (optional):
Layer 9
(a) 36
(b) Y-
( c ) /7. t?/?wr
(d) /J!
(e) ;^,-
(f) f7.,7-777
Layer 12
(a)
(b)
(c)
(d)
(e)
(f)
1 and 15, enter:
(1 to 5)
(0 to 100)
If the soil texture number of layer 1 is between 16 and 20, enter:
SCS runoff curve number: - (0 to 100)
If landfill is open, enter potential runoff fraction:
Surface area:
Slope of top liner/drain system: \J_
Distance from crest to drain in top liner/drain system:
Slope of second liner/drain system: >j
Distance from crest to drain in second liner/drain system:
Slope of third liner/drain system: ^j
Distance from crest to drain in third liner/drain system:
Slope of fourth liner/drain system: ___ZZZIZ
Distance from crest to drain in fourth liner/drain system:
(0 to 1)
Initial quantity of snow or ice water on surface (not asked if
program is to initialize the soil water): S?
square feet
percent
feet
percent
feet
percent
feet
percent
feet
inches
* If soil texture number is 19:
If soil texture number is 20:
(a) wilting point
(b) field capacity
(c) porosity
(d) saturated hydraulic
conductivity
vol/vol
vol/vol
vol/vol
era/sec
(a) wilting point
(b) field capacity
(c) porosity
(d) saturated hydraulic
conductivity
vol/voi
vol/vol
vol/vol
cm/sec
Figure 8-6. (Continued).
83
-------
area index were selected from the recommended values
for a fair stand of grass. A completed data form from the
user guide (5) for climatological data input is shown in
Figure 8-7.
A partial listing of the output giving all available options,
is shown in Figure 8-8. The options include daily, month-
ly, and annual water balances.
REFERENCES
1. U.S. EPA. 1989. Technical guidance document:
Final covers on hazardous waste landfills and sur-
face impoundments. EPA/530-SW-89-047.
2. U.S. EPA. 1988. U.S. EPA guide to technical resour-
ces for the design of land disposal facilities. EPA
Guidance Document: Final Covers on Hazardous
Waste Landfills and Surface Impoundments.
EPA/530-SW-88-047.
3. Schroeder, P.R., J.M. Morgan, T.M. Walski, and A.C.
Gibson. 1984a. Hydrologic Evaluation of Landfill Per-
formance (HELP) Model: Vol. I. User's Guide for
Version 1. EPA/530-SW-84-009. U.S. Environmental
Protection Agency, Washington, DC. 120 pp.
4. Schroeder, P.R., A.C. Gibson, and M.D. Smolen.
1984b. Hydrologic Evaluation of Landfill Performance
(HELP) Model: Vol. II. Documentation for Version 1.
EPA/530-SW-84-010. U.S. Environmental Protection
Agency, Washington, DC. 256 pp.
5. Schroeder, P.R., R.L. Peyton, and J. M. Sjostrom.
1988a. Hydrologic Evaluation of Landfill Performance
(HELP) Model: Vol. III. User's Guide for Version 2.
Internal Working Document. USAE Waterways Ex-
periment Station, Vicksburg, MS.
6. Schroeder, P.R., B.M. McEnroe, and R.L. Peyton.
1988b. Hydrologic Evaluation of Landfill Performance
(HELP) Model: Vol. IV. Documentation for Version 2.
Internal Working Document. USAE Waterways Ex-
periment Station, Vicksburg, MS.
7. Perrier, E.R. and A.C. Gibson. 1980. Hydrologic
simulation on solid waste disposal sites. EPA-SW-
868. U.S. Environmental Protection Agency, Cincin-
nati, OH. 111 pp.
8. Schroeder, P.R. and A.C. Gibson. 1982. Supporting
documentation 1or the Hydrologic Simulation Model
for Estimating Percolation at Solid Waste Disposal
Sites (HSSWDS). Draft Report. U.S. Environmental
Protection Agency, Cincinnati, OH. I53 pp.
9. Knisel, W.G., Editor. 1980. CREAMS, a field-scale
model for chemical runoff and erosion from agricul-
tural management systems. Vols. I, II, and III. USDA-
SEA-AR Conservation Research Report 26. 643 pp.
10. Williams, J.R., A.D. Nicks, and J.G. Arnold. 1985.
SWRRB, a simulator for water resources in rural
basins. Journal of Hydraulic Engineering. ASCE, Vol.
111, No. 6, pp. 970-986.
CLIMATOLOCICAL DATA INPUT
Default Precipitation Option
Normal Mean Monthly Temperatures in Degrees Fahrenheit (Optional)
Jan.
Feb.
Mar.
Apr.
May
Jun .
Jul.
Aug.
Sep.
Oct.
Nov.
Dec.
Haximuro Leaf Area Index"
Figure 8-7. Completed data form for climatological data.
84
-------
RCRA COVER SEMINAR
PHILADELPHIA, PENNSYLVANIA
29 MAY 90
FAIR GRASS
LAYER 1
VERTICAL PERCOLATION LAYER
THICKNESS - 24.00 INCHES
POROSITY - 0.3980 VOL/VOL
FIELD CAPACITY - 0.2443 VOL/VOL
WILTING POINT - 0.1361 VOL/VOL
INITIAL SOIL WATER CONTENT - 0.2739 VOL/VOL
SATURATED HYDRAULIC CONDUCTIVITY - 0.000360000005 CM/SEC
LAYER 2
LATERAL DRAINAGE LAYER
THICKNESS - 12.00 INCHES
POROSITY - 0.4170 VOL/VOL
FIELD CAPACITY - 0.0454 VOL/VOL
WILTING POINT - 0.0200 VOL/VOL
INITIAL SOIL WATER CONTENT - 0.3489 VOL/VOL
SATURATED HYDRAULIC CONDUCTIVITY - 0.009999999776 CM/SEC
SLOPE - 3.00 PERCENT
DRAINAGE LENGTH - 500.0 FEET
LAYER 3
BARRIER SOIL LINER
THICKNESS - 36.00 INCHES
POROSITY - 0.4300 VOL/VOL
FIELD CAPACITY - 0.3663 VOL/V°L
WILTING POINT - 0.2802 VOL/VOL
INITIAL SOIL WATER CONTENT - 0.4300 VOL/VOL
SATURATED HYDRAULIC CONDUCTIVITY - 0.000000100000 CM/SEC
Figure 8-8. Example output.
85
-------
LAYER 4
THICKNESS
POROSITY
FIELD CAPACITY
WILTING POINT
INITIAL SOIL WATER CONTENT
SATURATED HYDRAULIC CONDUCTIVITY
VERTICAL PERCOLATION LAYER
588.00 INCHES
0.5200 VOL/VOL
0.2942 VOL/VOL
0.1400 VOL/VOL
0.2840 VOL/VOL
0.000199999995 CM/SEC
LAYER 5
THICKNESS
POROSITY
FIELD CAPACITY
WILTING POINT
INITIAL SOIL WATER CONTENT
SATURATED HYDRAULIC CONDUCTIVITY
VERTICAL PERCOLATION LAYER
12.00 INCHES
0.5200 VOL/VOL
0.2942 VOL/VOL
0.1400 VOL/VOL
0.2852 VOL/VOL
0.000199999995 CM/SEC
LAYER 6
LATERAL DRAINAGE LAYER
THICKNESS
POROSITY
FIELD CAPACITY
WILTING POINT
INITIAL SOIL WATER CONTENT
SATURATED HYDRAULIC CONDUCTIVITY
SLOPE
DRAINAGE LENGTH
12.00 INCHES
0.4170 VOL/VOL
0.0454 VOL/VOL
0.0200 VOL/VOL
0.0454 VOL/VOL
0.009999999776 CM/SEC
5.00 PERCENT
50.0 FEET
LAYER 7
BARRIER SOIL LINER WITH FLEXIBLE
THICKNESS
POROSITY
FIELD CAPACITY
WILTING POINT
INITIAL SOIL WATER CONTENT
SATURATED HYDRAULIC CONDUCTIVITY
LINER LEAKAGE FRACTION
MEMBRANE LINER
6.00 INCHES
0.4170 VOL/VOL
0.0454 VOL/VOL
0.0200 VOL/VOL
0.4170 VOL/VOL
0.009999999776 CM/SEC
0.00005000
Figure 8-8. (Continued).
86
-------
LAYER 8
LATERAL DRAINAGE LAYER
THICKNESS - 12.00 INCHES
POROSITY - 0.4170 VOL/VOL
FIELD CAPACITY - 0.0454 VOL/VOL
WILTING POINT - 0.0200 VOL/VOL
INITIAL SOIL WATER CONTENT - 0.0478 VOL/VOL
SATURATED HYDRAULIC CONDUCTIVITY - 0.009999999776 CM/SEC
SLOPE - 5.00 PERCENT
DRAINAGE LENGTH - 50.0 FEET
LAYER 9
BARRIER SOIL LINER WITH FLEXIBLE MEMBRANE LINER
THICKNESS - 36.00 INCHES
POROSITY - 0.3777 VOL/VOL
FIELD CAPACITY - 0.2960 VOL/VOL
WILTING POINT - 0.2208 VOL/VOL
INITIAL SOIL WATER CONTENT - 0.3777 VOL/VOL
SATURATED HYDRAULIC CONDUCTIVITY - 0.000001650000 CM/SEC
LINER LEAKAGE FRACTION - 0.00005000
GENERAL SIMULATION DATA
SCS RUNOFF CURVE NUMBER - 85.56
TOTAL AREA OF COVER - 1000000. SQ FT
EVAPORATIVE ZONE DEPTH - 24.00 INCHES
UPPER LIMIT VEG. STORAGE - 9.5520 INCHES
INITIAL VEG. STORAGE - 6.5736 INCHES
INITIAL SNOW WATER CONTENT - 0.0000 INCHES
INITIAL TOTAL WATER STORAGE IN
SOIL AND WASTE LAYERS - 213.8724 INCHES
SOIL WATER CONTENT INITIALIZED BY USER.
CLIMATOLOGICAL DATA
DEFAULT RAINFALL WITH SYNTHETIC DAILY TEMPERATURES AND
SOLAR RADIATION FOR PHILADELPHIA PENNSYLVANIA
MAXIMUM LEAF AREA INDEX - 2.00
START OF GROWING SEASON (JULIAN DATE) - 115
END OF GROWING SEASON (JULIAN DATE) - 296
Figure 8-8. (Continued).
87
-------
NORMAL MEAN MONTHLY TEMPERATURES, DEGREES FAHRENHEIT
JAN/JUL FEB/AUG MAR/SEP APR/OCT MAY/NOV JUN/DEC
31.20
76.50
**********•:
DAY
1
2
3
4*
5
6
7
8
9
10
11
12
13
14*
15*
16*
360
361
362
363
364
365
33.10 41.80 52.90 62.80 71.60
75.30 68.20 56.50 45.80 35.50
VARIABLE 1: HEAD ON TOP OF LAYER 3
VARIABLE 2: PERCOLATION THROUGH LAYER 3
VARIABLE 3: PERCOLATION THROUGH LAYER 7
VARIABLE 4: PERCOLATION THROUGH LAYER 9
VARIABLE 5: LATERAL DRAINAGE FROM LAYER 2
VARIABLE 6: LATERAL DRAINAGE FROM LAYER 8
it************************************************************
DAILY OUTPUT FOR YEAR 74
RAIN RUNOFF
IN.
0.05
0.00
0.50
0.15
0.00
0.00
0.00
0.00
0.62
0.29
0.50
0.00
0.02
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.31
IN.
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.001
0.000
0.006
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
ET
IN.
.014
.043
.013
.011
.040
.047
.047
.043
.015
.012
.014
.034
.012
.000
.000
.000
.084
.066
.096
.090
.079
.016
VAR.
1
IN.
9.8
9.8
9.8
9.8
9.8
9.9
10.0
10.2
10.4
10.5
10.7
11.2
13.2
15.4
16.9
18.1
4.7
4.8
4.8
4.9
4.9
5.0
VAR.
2
IN.
0.0030
0.0042
0.0043
0.0043
0.0043
0.0043
0.0043
0 . 0044
0 . 0044
0 . 0044
0 . 0044
0 . 0044
0.0046
0.0047
0.0048
0.0050
0.0038
0.0038
0.0039
0.0039
0.0039
0.0039
VAR.
3
IN.
0.0037
0.0043
0.0043
0.0043
0.0043
0.0043
0.0043
0.0043
0.0043
0.0043
0.0043
0.0043
0.0042
0.0042
0.0042
0.0042
0.0041
0.0041
0.0041
0.0041
0.0041
0.0041
VAR.
4
IN.
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0 . 0000
0.0000
0 . 0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0 . 0000
0.0000
VAR.
5
IN.
0.012
0.017
0.017
0.017
0.017
0.017
0.017
0.018
0.018
0.018
0.018
0.019
0.020
0.020
0.020
0.020
0.010
0.010
0.010
0.010
0.010
0.010
VAR.
6
IN.
0.003
0.004
0.004
0.004
0.004
0.004
0.004
0.004
0.004
0.004
0.004
0.004
0.004
0.004
0.004
0.004
0.004
0.004
0.004
0.004
0.004
0.004
SOIL
WATER
IN/IN
0.2745
0.2727
0.2867
0.2909
0.2905
0.2894
0.2847
0.2790
0.2934
0.3059
0.3199
0.3133
0.3025
0.3016
0.3005
0.2995
0.2576
0.2530
0.2478
0.2427
0.2380
0.2452
Figure 8-8. (Continued).
88
-------
MONTHLY TOTALS FOR YEAR 74
JAN/JUL FEB/AUG MAR/SEP APR/OCT MAY/NOV JUN/DEC
PRECIPITATION (INCHES)
RUNOFF (INCHES)
EVAPOTRANSPIRATION
(INCHES)
LATERAL DRAINAGE FROM
LAYER 2 (INCHES)
PERCOLATION FROM
LAYER 3 (INCHES)
LATERAL DRAINAGE FROM
LAYER 6 (INCHES)
PERCOLATION FROM
LAYER 7 (INCHES)
LATERAL DRAINAGE FROM
LAYER 8 (INCHES)
PERCOLATION FROM
LAYER 9 (INCHES)
2.95
2.08
0.007
0.002
0.978
2.882
2.14
3.83
0.000
0.231
1.568
2.725
4.91
4.68
0.264
0.085
2.629
4.584
2.77
1.93
0.089
0.121
3.257
1.895
3.21
0.81
0.000
0.000
3.368
1.116
4.43
4.04
0.052
0.179
5.157
1.267
0.5736 0.5438 0.6033 0.5951 0.6026 0.5870
0.5887 0.5094 0.4260 0.3829 0.3204 0.2957
0.1476 0.1456 0.1605 0.1793 0.1683 0.1481
0.1387 0.1329 0.1241 0.1243 0.1169 0.1184
0.0006 0.0005 0.0006 0.0006 0.0006 0.0006
0.0006 0.0006 0.0006 0.0006 0.0006 0.0006
0.1310 0.1169 0.1283 0.1235 0.1275 0.1234
0.1276 0.1277 0.1237 0.1279 0.1237 0.1279
0.1320 0.1172 0.1284 0.1236 0.1274 0.1233
0.1275 0.1276 0.1236 0.1278 0.1237 0.1278
0.0001 0.0001 0.0001 0.0001 0.0001 0.0001
0.0001 0.0001 0.0001 0.0001 0.0001 0.0001
MONTHLY SUMMARIES FOR
AVG. DAILY HEAD ON
LAYER 3 (INCHES)
STD. DEV. OF DAILY HEAD
ON LAYER 3 (INCHES)
AVG. DAILY HEAD ON
LAYER 7 (INCHES)
STD. DEV. OF DAILY HEAD
ON LAYER 7 (INCHES)
15.12
11.24
4.19
0.74
0.00
0.00
0.00
0.00
19.00
9.33
0.20
0.51
0.00
0.00
0.00
0.00
DAILY HEADS
18.83
7.73
2.32
0.42
0.00
0.00
0.00
0.00
27.23
6.40
2.79
0.37
0.00
0.00
0.00
0.00
21.29
5.20
1.60
0.33
0.00
0.00
0.00
0.00
16.06
4.43
1.63
0.28
0.00
0.00
0.00
0.00
Figure 8-8. (Continued).
89
-------
AVG. DAILY HEAD ON 0.08
LAYER 9 (INCHES) 0.07
STD. DEV. OF DAILY HEAD 0.00
ON LAYER 9 (INCHES) 0.00
***********************************
£.£.£..£.^^^..£.^.£.£.£.££^£^.£^.^^^^^££.£^.,£,^££^.^^
ANNUAL TOTALS
PRECIPITATION
RUNOFF
EVAPOTRANSPIRATION
LATERAL DRAINAGE FROM LAYER 2
PERCOLATION FROM LAYER 3
LATERAL DRAINAGE FROM LAYER 6
PERCOLATION FROM LAYER 7
LATERAL DRAINAGE FROM LAYER 8
PERCOLATION FROM LAYER 9
CHANGE IN WATER STORAGE
SOIL WATER AT START OF YEAR
SOIL WATER AT END OF YEAR
SNOW WATER AT START OF YEAR
SNOW WATER AT END OF YEAR
ANNUAL WATER BUDGET BALANCE
0.07 0.
0.07 0.
0.00 0.
0.00 0.
**********
**********
FOR YEAR
(INCHES)
37.78
1.029
31.425
6.0285
1.7047
0.0070
1.5090
1.5097
0.0010
-2.221
213.87
211.65
0.00
0.00
0.00
07 0.07
07 0.07
00 0.00
00 0 . 00
*************!
*************!
74
(CU. FT.)
3148334.
85764.
2618786.
502377.
142062.
586.
125754.
125812.
85.
-185075.
17822700.
17637624.
0.
0.
-1.
0.07 0.07
0.07 0.07
0.00 0.00
0 . 00 0 . 00
*************
*************
PERCENT
100.00
2.72
83.18
15.96
4.51
0.02
3.99
4.00
0.00
-5.88
0.00
Figure 8-8. (Continued).
90
-------
AVERAGE MONTHLY VALUES IN INCHES FOR YEARS 74 THROUGH 78
PRECIPITATION
TOTALS
STD. DEVIATIONS
RUNOFF
TOTALS
STD. DEVIATIONS
EVAPOTRANSPIRATION
TOTALS
STD. DEVIATIONS
JAN/JUL
4.59
3.67
2.53
1.95
1.478
0.414
3.008
0.556
0.973
4.199
0.246
1.766
LATERAL DRAINAGE FROM LAYER
FEB/AUG
1.88
4.46
0.66
2.49
0.053
0.212
0.066
0.195
1.485
3.565
0.220
1.593
2
MAR/SEP
4.09
4.17
1.00
2.07
0.290
0.302
0.181
0.596
2.701
2.998
0.089
1.536
APR/OCT
3.03
2.76
1.51
1.21
0.154
0.047
0.216
0.050
3.019
2.027
0.331
0.836
MAY/NOV
3.85
2.68
2.03
2.63
0.156
0.384
0.254
0.706
4.110
1.361
1.020
0.430
JUN/DEC
4.50
3.99
2.17
1.78
0.170
0.339
0.244
0.370
4.833
1.015
1.223
0.195
TOTALS 0.4856 0.4957 0.5418 0.5476 0.5656 0.5414
0.5247 0.4530 0.3789 0.3728 0.3589 0.4268
STD. DEVIATIONS 0.1797 0.1440 0.1524 0.0954 0.0847 0.0987
0.1150 0.1025 0.0900 0.1297 0.1225 0.1686
PERCOLATION FROM LAYER 3
TOTALS 0.1548 0.1502 0.1639 0.1623 0.1567 0.1427
0.1344 0.1291 0.1209 0.1236 0.1204 0.1331
STD. DEVIATIONS 0.0391 0.0353 0.0286 0.0245 0.0182 0.0149
0.0082 0.0069 0.0060 0.0089 0.0101 0.0199
LATERAL DRAINAGE FROM LAYER 6
TOTALS 0.0006 0.0006 0.0006 0.0006 0.0006 0.0006
0.0006 0.0006 0.0006 0.0006 0.0006 0.0006
Figure 8-8. (Continued).
91
-------
STD. DEVIATIONS 0.0000
0.0000
PERCOLATION FROM LAYER 7
TOTALS 0.1301
0.1304
STD. DEVIATIONS 0.0015
0.0022
LATERAL DRAINAGE FROM LAYER
TOTALS
0.0000 0.0000 0.0000 0.0000 0.0000
0.0000 0.0000 0.0000 0.0000 0.0000
0.1181 0.1298 0.1257 0.1300 0.1260
0.1305 0.1264 0.1306 0.1264 0.1306
0.0023 0.0017 0.0018 0.0020 0.0021
0.0022 0.0021 0.0022 0.0021 0.0021
8
0.1302
0.1303
STD. DEVIATIONS 0.0017
0.0022
PERCOLATION FROM LAYER 9
TOTALS 0.0001
0.0001
STD. DEVIATIONS 0.0000
0.0000
0.1181
0.1304
0.0022
0.0022
0.1297
0.1263
0.1256
0.1305
0.1299
0.1263
0.1259
0.1305
0.0016 0.0018 0.0020 0.0020
0.0021 0.0022 0.0021 0.0022
0.0001 0.0001 0.0001 0.0001 0.0001
0.0001 0.0001 0.0001 0.0001 0.0001
0.0000 0.0000 0.0000 0.0000 0.0000
0.0000 0.0000 0.0000 0.0000 0.0000
AVERAGE ANNUAL TOTALS & (STD. DEVIATIONS) FOR YEARS 74 THROUGH 78
(INCHES)
PRECIPITATION
RUNOFF
EVAPOTRANSPIRATION
LATERAL DRAINAGE FROM
LAYER 2
PERCOLATION FROM LAYER 3
LATERAL DRAINAGE FROM
43.67 i
3.998 i
32.287 i
5.6928 i
1.6920 i
0.0073 i
( 7.930)
( 3.685)
( 2.428)
( 1.0786)
( 0.1508)
( 0.0002)
(CU. FT.)
3639167.
333190.
2690580.
474402.
140998.
604.
PERCENT
100.00
9.16
73.93
13.04
3.87
0.02
PERCOLATION FROM LAYER 7
Figure 8-8. (Continued).
1.5347 ( 0.0220)
127890.
3.51
92
-------
LATERAL DRAINAGE FROM
LAYER 8
PERCOLATION FROM LAYER 9
CHANGE IN WATER STORAGE
1.5338 ( 0.0214) 127814. 3.51
0.0010 ( 0.0000) 86. 0.00
0.150 ( 5.089) 12491. 0.34
PEAK DAILY VALUES FOR YEARS 74 THROUGH 78
PRECIPITATION
RUNOFF
LATERAL DRAINAGE FROM LAYER 2
PERCOLATION FROM LAYER 3
HEAD ON LAYER 3
LATERAL DRAINAGE FROM LAYER 6
PERCOLATION FROM LAYER 7
HEAD ON LAYER 7
LATERAL DRAINAGE FROM LAYER 8
PERCOLATION FROM LAYER 9
HEAD ON LAYER 9
SNOW WATER
(INCHES)
3.99
2.341
0.0209
0.0068
36.1
0.0000
0.0043
0.0
0.0045
0.0000
0.1
4.09
(CU. FT.)
332500.0
195074.9
1744.7
567.3
2.5
359.0
371.3
0.2
340770.0
MAXIMUM VEG. SOIL WATER (VOL/VOL)
MINIMUM VEG. SOIL WATER (VOL/VOL)
0.3980
0.1359
Figure 8-8. (Continued).
93
-------
FINAL WATER STORAGE AT END OF YEAR 78
LAYER
1
2
3
4
5
6
7
8
9
SNOW WATER
(INCHES)
6.57
4.19
15.48
167.74
3.42
0.55
2.50
0.57
13.60
0.00
(VOL/VOL)
0.2739
0.3488
0.4300
0.2853
0.2853
0.0454
0.4170
0.0478
0.3777
Figure 8-8. (Continued).
94
-------
11. USDA, Soil Conservation Service. 1972. Section 4,
Hydrology. In: National Engineering Handbook, U.S.
Government Printing Office, Washington, DC. 631
PP-
12. Horton, R.E. 1919. Rainfall Interception. Monthly
Weather Review, U.S. Weather Bureau. Vol. 47, No.
9, pp. 603-623.
13. Ritchie, J.T. 1972. A model for predicting evaporation
from a row crop with incomplete cover. Water
Resources Research. Vol. 8, No. 5, pp. 1204-1213.
14. Brooks, R.H. and AT. Corey. 1964. Hydraulic proper-
ties of porous media. Hydrology Paper No. 3,
Colorado State University. 27 pp.
15. Campbell, G.S. 1974. A simple method for determin-
ing unsaturated hydraulic conductivity from moisture
retention data. Soil Science. Vol. 117, No. 6, pp.
311-314.
16. Brown, K.W., J.C. Thomas, R.L Lytton, P. Jayawik-
rama, and S.C. Bahrt. 1987. Quantification of Leak
Rates Through Holes in Landfill Liners. EPA/600/S2-
87-062. U.S. Environmental Protection Agency, Cin-
cinnati, OH. 151 pp.
17. Giroud, J.P. and R. Bonaparte. 1989a. Leakage
through liners constructed with geomembranes,
Part I: Geomembrane liners. Geotextiles and
Geomembranes. Vol. 8, No. 1, pp. 27-67.
18. Giroud, J.P. and R. Bonaparte. 1989b. Leakage
through liners constructed with geomembranes,
Part II: Composite liners. Geotextiles and
Geomembranes. Vol. 8, No. 2, pp. 71 -111.
19. Giroud, J.P., A. Khatami, and K. Badu-Tweneboah.
1989c. Evaluation of the rate of leakage through
composite liners. Geotextiles and Geomembranes.
Vol. 8, No. 4, pp. 337-340.
20. Richardson, C.W. and D.A. Wright. 1984. WGEN: a
model for generating daily weather variables. ARS-8,
Agricultural Research Service, USDA. 83 pp.
95
-------
CHAPTER 9
SENSITIVITY ANALYSIS OF HELP MODEL PARAMETERS
INTRODUCTION
This chapter examines the sensitivity of landfill water
balance to numerous landfill design variables using the
Hydrologic Evaluation of Landfill Performance (HELP)
model. This information is useful in a variety of ways. It
can aid the design engineer in selecting preliminary
design alternatives for municipal or hazardous waste
landfills. It can serve as a basis for regulatory agencies
to establish and evaluate technical guidelines. It can
also provide additional insight on the importance and in-
teraction of specific design variables on the water
balance. Finally, it can assist in evaluating the suitability
of methodologies used in the computer model. The
analyses include examination of both cover systems and
lateral drainage/liner systems (1). The complete list of
design characteristics examined is given in Table 9-1.
The analysis of landfill cover design is divided into two
parts. First, water balance results are compared for dif-
ferent general design conditions such as climate (loca-
tion), topsoil and vegetative characteristics, and cover
Table 9-1. Parameters Selected for Sensitivity Analysis
Typical Cover Systems
Quantity of vegetation
Cover soil thickness
Topsoil type
Use of lateral drainage layer
Geographical location or climate
Vegetative Layer
SCS runoff curve number
Evaporative depth
Drainable porosity
Plant available water
Municipal vs. hazardous waste cover design
Analysis of Percolation and Drainage Design
Hydraulic conductivity of barrier soil layer
Hydraulic conductivity of lateral drainage layer
Geomembrane leakage factor
Liner type (clay, geomembrane, or composite)
Slope of lateral drainage layer
Drainage length
Double liner system design
design. Then, the effects resulting from changes in
specific characteristics of the vegetative layer, such as
runoff curve number, evaporative depth, and moisture
retention properties, are examined. The water balance
components examined in this chapter are surface runoff,
evapotranspiration, lateral subsurface drainage to collec-
tion systems, and vertical percolation through the soil
liner.
The analysis of liner systems examines the effects of
slope, drain spacing, saturated hydraulic conductivity,
and geomembrane leakage characteristics on leachate
collection and leakage through liners. Two types of verti-
cal inflows to the drain layer are considered. First, an in-
flow rate of 127 cm/yr (50 in./yr) was used to represent
infiltration at an open landfill. This inflow was distributed
in time according to actual rainfall patterns at Shreveport,
Louisiana. Second, an inflow rate of 20 cm/yr (8 in./yr)
uniformly distributed in time was used to represent in-
filtration at a covered landfill.
In the discussion that follows, the effects of the saturated
hydraulic conductivities of the drain layer and liner are
first investigated by holding the slope and drainage
length constant. Then, the slope and drainage length are
examined by holding the hydraulic conductivities con-
stant. In all cases, the thickness of the lateral drainage
layer was greater than the maximum head, and the thick-
ness of the soil liner was 61 cm (24 in.).
COMPARISON OF TYPICAL COVER SYSTEMS
Design Parameters
Three locations were studied to determine the effect of
various climatological regimes on cover performance—
Santa Maria, California; Schenectady, New York; and
Shreveport, Louisiana. These locations represent a wide
range in levels of precipitation, temperature, and solar
radiation as summarized in Table 9-2. Default values for
precipitation, temperature, solar radiation, and leaf area
index are stored in the HELP model for each site and
were used for the sensitivity analysis simulations. The
period of record stored in the HELP model for daily
precipitation is 1974 through 1978.
Two cover designs were examined as shown in Figure 9-1.
One is typical of some newer landfills where 0.61 m (2 ft)
97
-------
Table 9-2. Climatological Regimes
Location
Climatological
Variable
Santa Maria, CA Schenectady, NY
Shreveport, LA
Precipitation1
Mean annual (in.)
Mean winter
(Nov-Apr) (in.)
Mean summer
(May-Oct) (in.)
Temperature
Mean annual (°F)
Mean Jan (°F)
Mean July (°F)
Days with minimum
below 32°F
Solar radiation
Mean daily (langleys)
14
12
2
57
51
62
24
450
48
19
29
49
23
73
129
290
44
22
22
66
47
83
37
410
1These mean values are for the period simulated by the HELP model in this section, 1974-1978.
of topsoil overlies a 0.31-m (1-ft) thick lateral drainage
layer having a saturated hydraulic conductivity of 3 x 10~2
cm/sec, a slope of 0.01 m/m (0.03 ft/ft) and a maximum
drainage length of 61 m (200 ft). The drainage layer is
underlain by a 0.61-m (2-ft) thick soil liner having a
saturated hydraulic conductivity of 1 x 10~7 cm/sec. The
other design is typical of older municipal sanitary landfills
where a topsoil layer overlies a 0.61-m (2-ft) thick soil
liner having a saturated hydraulic conductivity of 1 x 10~6
cm/sec.
Two types of topsoil were considered in the cover
designs: sandy loam and silty, clayey loam. The sandy
loam characteristics were those of the HELP model
default soil texture 6, which represents Unified Soil Clas-
sification System (USCS) soil class SM and U.S. Depart-
ment of Agriculture (USDA) soil class SL. The silty,
clayey loam characteristics were those of the HELP
model default soil texture 12, which represents USCS
soil class CL and USDA soil class SICL. The topsoil-
type designation was used to select soil porosity, field
capacity, wilting point, and hydraulic conductivity, be-
sides influencing the selection of the runoff curve num-
ber. In addition to two types of topsoil, two thickness of
topsoil were examined—46 cm (18 in.) and 91 cm (36 in.).
The vegetative cover was designated as being either a
good stand of grass or a poor stand of grass. This selec-
tion dictated the values for leaf area index, evaporative
depth, and runoff curve number, and influenced the value
used for the saturated hydraulic conductivity of the top-
soil. For a given vegetative cover and topsoil material,
the runoff curve number was obtained from the HELP
Model User's Guide (2). These numbers were 60 for
good grass on sandy loam; 80 for poor grass on sandy
loam; 81 for good grass on silty, clayey loam; and 92 for
poor grass on silty, clayey loam. These curve numbers
are in agreement with values obtained from Section 4,
Hydrology, National Engineering Handbook (3). The
depth of the evaporative zone was chosen as 18 cm (7
in.) for poor grass and 36 cm (14 in.) for good grass.
Results
Figures 9-2 and 9-3 summarize the results obtained in
the general sensitivity analysis performed on the cover
systems, respectively, with and without lateral drainage.
The height of each bar segment represents the cor-
responding mean annual value of water balance com-
ponent in inches which is given next to each bar
segment. The results provide a comparison of the effects
of varying quantity of vegetation, cover design, topsoil
type, topsoil thickness, and Climatological regime.
Effects of Vegetation
Two levels of vegetation were examined—a poor stand of
grass and a good stand of grass; the latter represents
three times the quantity of vegetation as that of the
former. Table 9-3 presents the water balance results for
both cover systems at all three sites as a function of level
of vegetation. The results are given in units of percent of
the precipitation during the simulation period.
Vegetation reduces surface runoff and increases
evapotranspiration. Evapotranspiration is greater be-
cause the plant demand for moisture and a greater quan-
tity of water is available for evapotranspiration due to
greater infiltration and a greater evaporative zone. Runoff
is less because vegetation increases the minimum in-
filtration rate, drying rate, interception, and surface rough-
ness, which results in a decrease in the runoff curve
98
-------
VEGETATION
LATERAL DRAINAGE
18" or 36"
12"
24"
a) Three Layer Landfill Cover Design
VEGETATION
18" or 36"
24'
b) Two Layer Landfill Cover Design
Figure 9-1. Cover designs for sensitivity analysis.
number. The influence of surface vegetation on the
volume of lateral drainage and percolation or leakage
from the cover is varied. However, the quantity of
vegetation tends to have very little effect on the percola-
tion or leakage through the cover system. For the cover
with lateral drainage, the increase in infiltration with good
grass was greater than the increase in evapotranspira-
tion, resulting in a larger volume of lateral drainage and a
negligible change in percolation. For the cover without
lateral drainage, the increase in infiltration yielded high
heads or depths of saturation above the liner that per-
mitted greater evapotranspiration by maintaining higher
moisture contents in the evaporative zone. Consequent-
ly, the increase in evapotranspiration was greater than
the increase in infiltration. This resulted in a trend toward
a small decrease in percolation for a higher level of
vegetation. The opposite trend may occur for vegetative
layers having lower saturated hydraulic conductivities
and higher plant available water capacities. The results
were similar at all three sites despite quite different
climates. In summary, vegetation decreases runoff and
increases evapotranspiration but tends to have little ef-
fect on the water balance. The magnitude of the effects
is design dependent and to a lesser degree climate de-
pendent. The main function of vegetation is to control
erosion.
Effects of Topsoil Thickness
Two topsoil thicknesses were examined—46 cm (18 in.)
and 91 cm (36 in.). Table 9-4 presents the water balance
results for the two-layer cover system as a function of
topsoil thickness at all three sites. The results are given
in units of percent of the precipitation during the simula-
tion period. The cover system with lateral drainage was
not used in this analysis because lateral drainage would
negate the effects by preventing or minimizing the in-
99
-------
60 -
55 -
~ 50 -
UJ
^
= 40 -
UJ
3
£ 35 -
^ 30 -
I 25 -
| 20 -
UJ
* IS -
10 -
5 -
0 -
(
1
iy
A
I
QQ
PQ
SL
RUNOFF
EVAPOTRANSPIRATION
LATERAL DRAINAGE
PERCOLATION
GOOD GRASS
POOR GRASS
18- OF SANDY LOAM
SICL 18" OF SILTY
T10.0
7.8
6.2
i -0.0
3G
SL
CLAYEY LOAM
0.4 n1.1 n3.1
7.4 10.4 8.8
5.8 2. S 2.2
0.8 0.3 0.3
0.1
23.3
19.3
1.4
2.0
22.
18.
1 i
PG GG PG GG PG
SICL SL
9
0
3.8
32.8
6.6
1.0
8.8
28.
6.0
0.9
GQ PG
SICL
SANTA MARIA. CA SHREVEPORT, LA
4
0.0
24.7
22.1
1 ")
' 1 *\ -. . W
1.1
24.
21.
1.2
1
GG PG
SL
3
4
3.8
31.1
12.3
1.0
18.4
27.9
9.8
1.0
GG PG
SICL
6CHENECTADY. NY
Figure 9-2. Bar graph for three-layer cover design showing effect of surface vegetation, topsoil type, and location.
55 -
_ 53 -
CO
£ 45 -
o
- 40 -
UJ
3 35 -
-i 30 -
1 25 -
* 20 -
UJ
2 IS -
10 -
5 -
-*
(
T1.3
' 8.3
4.9
3G
'//'
'$\$f
GG
PG
18
36
RUNOFF
EVAPOTRANSPIRATION
PERCOLATION
GOOD GRASS
POOR GRASS
18- OF SANDY LOAM
36" OF SANDY LOAM
/
<
j
1.8
7.5
5.3
\
PG G
18
ijO.4 D0.8
7.8 ' 7.S
8.1 8.1
G PG G
1.0
29.4
13.8
G P
3.3
25.0
15.8
G
36 18
G
;zs.i
18.9
G P
2.0
23.
3
18.9
G
36
SANTA MARIA, CA SHREVEPORT. LA
G
4.6
29.4
14.0
G P
8.5
25.
a
15.8
3
18
G
1.7
28.3
19.9
3 P
]2.7
.
24.7
I 20.8
3
36
SCHENECTADY, NY
Figure 9-3. Bar graph for two-layer cover design showing effect of topsoil depth, surface vegetation, and location.
100
-------
Table 9-3. Effects of Climate and Vegetation
81 cm (36 in.) of Sandy Loam Topsoil
61 cm (24 in.) of 1 x 10"6 cm/sec Clay Liner
Locations
Poor grass
Runoff
Evapotranspiration
Percolation
Good grass
Runoff
Evapotranspiration
Percolation
CA
5.6
51.8
42.6
3.1
55.0
42.9
LA
(Percent of Precipitation)
4.6
53.0
42.4
0.2
57.2
42.6
NY
5.5
52.1
42.4
3.5
55.3
41.2
46 cm (18 in.) of Sandy Loam Topsoil
31 cm (12 in.) of 0.03 cm/sec Sand with 61 m (200 ft) Drain Length at 3% Slope
,-7.
61 cm (24 in.) of 1x10 cm/sec Clay Liner
Locations
Poor grass
Runoff
Evapotranspiration
Lateral drainage
Percolation
Good grass
Runoff
Evapotranspiration
Lateral drainage
Percolation
CA
3.0
51.6
41.2
4.2
0.0
52.6
43.2
4.2
LA
(Percent of Precipitation)
4.4
51.9
40.6
3.1
0.2
53.0
43.7
3.1
NY
2.2
50.3
44.0
2.5
0.0
51.0
45.5
2.5
Table 9-4. Effects of Climate and Topsoil Thickness
Sandy Loam Topsoil with a Poor Stand of Grass
61 cm (24 in.) of 1 x 10"6 cm/sec Clay Liner
46cm (18 in.) of topsoil
Runoff
Evapotranspiration
Percolation
91 cm (36 in.) of topsoil
Runoff
Evapotranspiration
Percolation
Locations
CA
11.2
51.9
36.9
5.6
51.8
42.6
LA
(Percent of Precipitation)
7.5
56.9
35.6
4.6
53.0
42.4
NY
13.4
54.5
32.1
5.5
52.1
42.4
101
-------
trusion of the saturated zone above the liner into the
evaporative zone.
Significant differences existed between the 46- and 91-
cm (18- and 36-in.) topsoil depth simulations in the ab-
sence of lateral drainage. The effects were similar at all
three sites. Runoff and evapotranspiration were greater
for the shallower depth to the liner, indicating that the
head above the barrier soil layer maintained higher mois-
ture contents in the evaporative zone. The percolation
was subsequently less than the cases with greater top-
soil thickness. The 91-cm (36-in.) depth to the liner per-
mits larger heads and longer sustaining heads since a
greater thickness of material below the evaporative zone
is free from abstraction of water by evapotranspiration.
The larger heads provide a greater pressure gradient to
increase the leakage rate through the cover system.
In general, the effects of topsoil thickness vary greatly as
the thickness increases from several inches to several
feet. Throughout the transition, the quantity of runoff
should continue to decrease until the depth to the liner
becomes sufficiently great so as to prevent the zone of
saturation to ever climb into the evaporative zone.
Similarly, the percolation through the liner should con-
tinue to increase until there is no interaction between the
saturation zone and the evaporative zone. The
evapotranspiration is expected to increase initially as the
available storage in the evaporative zone increases, i.e.,
until the depth to the liner equals the maximum depth
that evapotranspiration can reach. At greater depths the
evapotranspiration should continue to decrease until the
depth to the liner is sufficient to prevent any further inter-
actions between the evaporative and saturation zones.
While percolation increases with topsoil thickness given
identical properties for all layers in the cover system,
adequate thickness must be provided in a design to en-
sure the integrity of the cover system. A small topsoil
thickness would not provide adequate water storage to
support vegetation, maintain soil stability, and control
erosion. Similarly, a shallow depth to the liner would
promote desiccation or freezing of the liner, which may
greatly increase its permeability and, therefore, the per-
colation.
Effects of Topsoil Type
Two topsoil types were examined—sandy loam and silty,
clayey loam. Table 9-5 presents the water balance
results for the three-layer cover system as a function of
topsoil type at all three sites. The results are given in
units of percent of the precipitation during the simulation
period. The cover system without lateral drainage was
not used in this analysis because the intrusion of the
saturated zone above the liner into the evaporative zone
would decrease the magnitude of the effects.
The results show that the clayey topsoil significantly in-
creased both runoff and evapotranspiration, which in turn
greatly decreased lateral drainage and percolation. The
results were similar at all three sites. Runoff increased
from about 3 percent to 20 percent of the precipitation,
due primarily to the larger runoff curve number selected
for the clayey soil based on its lower minimum infiltration
rate. Evapotranspiration increased approximately from
51 percent to 61 percent of precipitation, due to the lower
hydraulic conductivity of the clayey soil and, more impor-
tantly, the larger plant available water capacity (field
capacity minus wilting point). The lower hydraulic con-
ductivity of the clayey soil slowed the drainage rate,
maintaining moisture contents above field capacity for
longer periods of time and allowing greater
evapotranspiration. The larger plant available water
capacity of the clayey soil provided a larger moisture
reservoir available for evapotranspiration after gravity
drainage ceased. The lateral drainage was reduced from
about 42 percent to 16 percent of the precipitation and
Table 9-5. Effects of Climate and Topsoil Types
46 cm (18 in.) of Topsoil with a Poor Stand of Grass
31 cm (12 in.) of 0.03 cm/sec Sand with 61 m (200 ft) Drain Length at 3% Slope
,-7.
61 cm (24 in.) of 1 x 10 cm/sec Clay Liner
Sandy loam
Runoff
Evapotranspiration
Lateral drainage
Percolation
Silty, clayey loam
Runoff
Evapotranspiration
Lateral drainage
Percolation
Locations
CA LA NY
(Percent of Precipitation)
3.0 4.4 2.2
51.6 51.9 50.3
41.2 40.6 44.0
4.2 3.1 2.5
21.6 22.3 19.2
61.2 64.4 58.6
15.0 11.3 20.3
2.2 2.0 1.9
102
-------
the percolation was reduced from about 3 percent to 2
percent of precipitation.
Use of Lateral Drainage Layer
Direct comparison of the use of a lateral drainage layer
was not made since different liner systems were used in
the two cover designs. The impact of the use of a lateral
drainage layer was explained briefly above. In general,
the use of a lateral drainage layer would be expected to
decrease the height of the saturation zone above the
liner by draining some of the infiltrated water from the
cover system. As such, percolation through clay liners
would decrease slightly. In addition, runoff and
evapotranspiration also would tend to decrease but the
magnitude of the change would be design dependent.
Topsoil thickness, topsoil type, vegetation, and climate
would have impacts.
Effects of Climate
The effects of climate were examined in each of the pre-
vious sections. As shown in Figures 9-2 and 9-3, climate
affects the absolute magnitude, in inches, of the water
budget components. However, Tables 9-3, 9-4, and 9-5
show that climate has a much smaller effect on the rela-
tive magnitude of the water budget components in terms
of percent of the precipitation. The relative proportions of
the water budget components are primarily design de-
pendent while the magnitudes are strongly dependent on
the magnitude of the precipitation.
The effect of temperature and solar radiation can be
determined by comparing the results for the Louisiana
and New York sites. These two sites have similar annual
rainfall, although the New York site had somewhat higher
annual and summer rainfall. The higher temperature and
solar radiation in Louisiana produced about an inch or
two more evapotranspiration despite the larger quantity
of rainfall in New York. Consequently, the lateral
drainage tended to be slightly less at the Louisiana site.
However, these differences are much smaller than the
differences caused by changes in designs.
Vegetative Layer Properties
The effects of vegetative layer properties on the water
balance of two cover systems are presented below. The
vegetative layer properties examined are runoff curve
number, evaporative depth, drainable porosity, and plant
available water capacity. Vegetated cover designs with
and without lateral drainage were used in the analyses;
the vegetation was assumed to be a fair stand of grass.
The thickness of the vegetative layer was 46 cm (18 in.)
in both designs. The simulations were performed using
climatic data for Santa Maria, California, and Shreveport,
Louisiana, and topsoil properties typical of sandy loam
and silty, clayey loam. Tables 9-6 and 9-7 summarize
the parameter combinations examined under this part of
the sensitivity analysis study and present the results of
the simulations as percentage of precipitation.
Effects of SCS Runoff Curve Number
The SCS runoff curve number was varied from 65 to 90
for the sandy loam and from 75 to 95 for the silty, clayey
loam. The range of curve number was selected to in-
clude values representative of the entire range of pos-
sible slopes and land management practices used at
landfills. The depth of the evaporative zone was 25 cm
(10 in.) in all cases. Simulations for the three-layer cover
design were performed for both soil types, whereas
simulations for the two-layer cover design were per-
formed only for sandy loam. The results are presented in
Table 9-6.
An increase in runoff curve number produced an increase
in runoff and a decrease in evapotranspiration, lateral
drainage, and percolation. The percent increase in runoff
was less for the two-layer cover design than for the three-
layer cover design. This result was due to the higher
average moisture content in the topsoil layer of the two-
layer design caused by the restriction to vertical flow im-
posed by the soil liner in the absence of lateral drainage.
This limited the infiltration capacity of the topsoil, causing
more frequent saturation of the topsoil and, therefore,
more runoff. Thus, runoff volume at low curve numbers
was higher for the two-layer cover compared to the three-
layer cover. This effect was not as great at high curve
numbers because infiltration for both designs was sig-
nificantly reduced by the curve number itself rather than
saturated conditions.
The effects of location or climate on runoff are difficult to
discern from the results; however, results in terms of per-
cent of the precipitation did not differ greatly between the
two sites. For example, in comparing runoff from Santa
Maria and Shreveport, a smaller percentage of precipita-
tion could be expected to drain from the surface as runoff
in Santa Maria due to the higher evaporative demand
combined with lower total precipitation and longer periods
of time between storms. This effect is seen in the data
for the three-layer design, but the difference is not as
large as may have been expected. Only small differen-
ces occur largely because the majority of the rainfall at
Santa Maria occurs during the winter when the evapora-
tive demand is the lowest. In addition, several unusually
large storms occurred at Santa Maria that yielded un-
usually large runoff. However, for simulations of the two-
layer design with low curve numbers, the influence of the
two large storms in Santa Maria caused the runoff per-
centage to exceed that in Shreveport. This would not be
the case if the two storms were excluded.
Summarizing the curve number effects, increasing the
curve number directly causes an increase in runoff and a
decrease in infiltration. The majority of the decrease in
infiltration is reflected as decreases in lateral drainage
and evapotranspiration. The decrease in leakage
through the cover system is generally small. Changes in
slope, vegetation, and land management practices yield
103
-------
only small changes in runoff for soil types and conditions
with curve numbers below 75. The climate, design, and
topsoil characteristics affect the volume of runoff for a
given curve number. The nature of the effects is closely
tied to the potential for evapotranspiration, vertical
drainage from the topsoil, and lateral drainage.
Effects of Evaporative Depth
Evaporative depth as defined by its use in the HELP
model is the thickness of the evapotranspiration zone,
the maximum depth from which water can be extracted to
satisfy evapotranspiration demand. This depth is a func-
tion of soil properties, vegetation, climate, and design.
The evaporative depth was varied from 10 to 46 cm (4 to
18 in.) for both sandy loam and silty, clayey loam. The
runoff curve number was 75 for the sandy loam and 85
for the silty, clayey loam. Simulations for the three-layer
cover design were performed for both soil types, whereas
simulations for the two-layer cover design were per-
formed only for sandy loam. The results are presented in
Table 9-6.
Evapotranspiration increased with increasing evaporative
depth while lateral drainage and percolation decreased;
the effect on runoff varied. The interrelationship between
these variables is complex and depends on many fac-
tors. The increase in evaporative depth allows
evapotranspiration to deplete soil moisture from greater
depths, generally increasing the total volume of
evapotranspiration. However, since the total
evapotranspiration demand remains constant, a smaller
volume of water depletion occurs per unit depth. Conse-
quently, the average moisture content throughout the
evaporative zone would be higher, resulting in a higher
runoff curve number and, therefore, larger runoff.
However, when the time period between storms is suffi-
ciently long, evapotranspiration demand is able to
deplete soil moisture to equal levels with either small or
large evaporative depths. In this case, runoff volume
could decrease with increasing evaporative depth since
antecedent moisture conditions would remain the same
and the increased storage volume in the deeper evapora-
tive zone would increase the infiltration capacity.
The effect of evaporative depth on the volume of
lateral drainage and percolation is directly related to the
composite effect on evapotranspiration and runoff. In the
examples chosen for Table 9-6, the increase in
evapotranspiration with increased evaporative depth was
greater than any increase in infiltration; therefore, lateral
drainage and percolation always decreased.
An increase in evaporative depth caused an increase in
infiltration for the two-layer cover compared to a slight
decrease for the three-layer cover. This difference re-
lates to the different mechanisms controlling infiltration in
these two cases. For the two-layer cover, the hydraulic
conductivity of the clay liner was much less than the
sandy loam topsoil. Therefore, infiltration tended to
saturate the topsoil layer, and the total volume of infiltra-
tion was dependent primarily on the volume of storage
available in this layer. A larger evaporative depth in-
creased the potential for a larger volume of available
storage and thus for more infiltration. For the three-layer
cover, the lateral drainage layer generally maintained a
free drainage condition at the topsoil/lateral drainage
layer interface. Infiltration was then controlled primarily
by the hydraulic conductivity of the topsoil and the avail-
able storage in the top segment of the subprofile. As ex-
plained above, this condition could result in either an
increase or decrease in infiltration with an increase in
evaporative depth.
Summarizing the effects of evaporative depth, an in-
crease in evaporative depth produces an increase in
evapotranspiration and, therefore, generally a decrease
in lateral drainage and percolation. The effects on runoff
are mixed but typically very small. The size of the chan-
ges are difficult to predict because the effects of evapora-
tive depth changes are indirect. Changing the
evaporative depth changes the potential storage in the
potential storage in the evaporative zone that may not
significantly change the net evapotranspiration. As
evidence of this, the change in evapotranspiration is very
small when the evaporative depth is increased beyond 46
cm (18 in.). In addition, the topsoil characteristics,
climate, and design affect the response to a change in
evaporative depth.
Effects of Drainable Porosity
Drainable porosity is defined as the difference between
porosity and field capacity; that is, the amount of water
that could be vertically drained from a saturated soil by
gravity forces alone. Values ranged from 0.254 to 0.686
cm/cm (0.100 to 0*270 in./in.) in this study. These values
represent the volume of moisture storage capacity in ex-
cess of field capacity, divided by the bulk volume of soil
including voids. Values for field capacity and wilting point
remained constant at 0.668 and 0.338 cm/cm (0.263 and
0.133 in./in.), respectively. Only sandy loam soil was
considered. The evaporative depth was 25 cm (10 in.),
and the SCS curve number was 75. Both two- and three-
layer cover designs were simulated. The results are
presented in Table 9-7.
An increase in drainable porosity increases the moisture
storage volume above field capacity and decreases un-
saturated hydraulic conductivity for a given moisture con-
tent given a constant saturated hydraulic conductivity.
Therefore, more water can infiltrate and be made avail-
able for evapotranspiration during vertical drainage. This
increases the volume of evapotranspiration and
decreases the volume of lateral drainage and percolation
as shown in Table 9-7. However, the effect of increased
drainable porosity on runoff is varied. For the three-layer
cover, runoff decreased slightly at Santa Maria and in-
creased slightly at Shreveport. For the two-layer cover,
104
-------
Table 9-6. Effects of Evaporative Depth and Runoff Curve Number
Description1
Average Annual Volume (Percent Precipitation)
Three-Layer Cover Design
Two-Layer Cover Design
Site
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
LA
LA
LA
LA
LA
LA
LA
LA
LA
LA
LA
LA
Soil
Type
SL
SL
SL
SICL
SICL
SICL
SL
SL
SL
SICL
SICL
SICL
SL
SL
SL
SICL
SICL
SICL
SL
SL
SL
SICL
SICL
SICL
Evap.
Depth
(in.)
10
10
10
10
10
10
4
10
18
4
10
18
10
10
10
10
10
10
4
10
18
4
10
18
SCS
Curve
Number
65
80
90
75
85
95
75
75
75
85
85
85
65
80
90
75
85
95
75
75
75
85
85
85
Runoff
0.1
2.6
11.3
5.5
12.7
34.4
1.1
1.1
1.3
12.6
12.7
12.0
0.5
4.2
15.3
5.8
13.5
36.5
2.0
2.1
2.3
12.4
13.5
14.3
ET3
52.7
51.9
49.5
70.8
67.6
57.3
41.3
52.4
61.9
53.3
67.6
77.0
52.1
50.9
47.1
71.2
69.6
59.0
38.8
51.6
62.4
55.6
68.1
75.8
Lat.4
Drng.
43.6
41.9
35.9
22.1
18.0
6.4
53.3
42.9
34.1
30.5
18.0
11.2
44.1
41.6
34.5
20.3
14.5
3.0
55.7
43.0
32.0
28.8
14.4
8.1
Liner
Perc.
4.2
4.2
4.1
2.2
2.2
1.6
4.5
4.2
3.9
3.7
2.2
1.2
3.1
3.1
3.0
2.3
2.2
1.4
3.2
3.1
3.0
2.0
2.1
1.2
Runoff
7.1
8.7
14.4
8.9
7.8
6.9
2.0
5.1
15.6
8.2
3.3
3.0
ET3
53.8
53.0
50.4
42.9
53.4
63.8
57.9
55.9
49.1
45.1
57.0
66.5
Liner6
Perc.
39.9
39.1
36.0
48.5
39.6
30.6
39.4
38.3
34.8
45.2
39.0
30.2
1CA = Santa Maria, CA; LA = Shreveport, LA; SL = sandy loam (HELP model default texture 6); SICL = silty, clayey loam (HELP
model default texture 12). Fair grass and 46-cm (18-in.) topsoil layer was used for all cases.
2Change in storage is not included in this table; therefore, the water balance components shown do not always add up to 100.0
percent.
3ET = evapotranspiration.
4Lateral drainage from a 31-cm (12-in.) layer having a slope of 3 percent, a drainage length of 61 m (200 ft), and a hydraulic
conductivity of 3 x 10"2 cm/sec.
5Percolation through 61-cm (24-in.) liner having a hydraulic conductivity of 10~7 cm/sec.
6Percolation through 61-cm (24-in.) liner having a hydraulic conductivity of 10~6 cm/sec.
runoff decreased significantly at both locations since the
relative soil moisture is lower and the available storage is
greater. An increase in drainable porosity reduces the
head or depth of saturation resulting from a fixed quantity
of infiltration. This decreases the lateral drainage while
having only small effects on percolation. The design and
climate affects the magnitudes of the changes in the
water budget components.
Effects of Plant Available Water Capacity
Plant available water capacity is defined as the difference
between field capacity and wilting point, or the amount of
water available for plant uptake after vertical drainage by
gravity has ceased. Values ranged from 0.178 to 0.508
cm/cm (0.070 to 0.200 in./in.) in this analysis. These
values represent the volume of potential moisture storage
between wilting point and field capacity, divided by the
105
-------
Table 9-7. Effects of Drainable Porosity and Plant Available Water Capacity
Description1
Site
CA
CA
CA
CA
CA
CA
LA
LA
LA
LA
LA
LA
DP
0.18
0.18
0.18
0.10
0.18
0.27
0.18
0.18
0.18
0.10
0.18
0.27
PAWC
0.07
0.13
0.20
0.13
0.13
0.13
0.07
0.13
0.20
0.13
0.13
0.13
Average Annual Volume (Percent Precipitation)2
Three-Layer Cover Design Two-Layer Cover Design
Runoff
1.07
1.14
1.30
1.17
1.14
1.1
2.08
2.15
2.26
2.10
2.15
2.2
ET3
48.51
52.54
56.43
48.87
52.53
55.8
47.38
51.74
55.68
46.93
51.74
55.7
Lat.4
Drng.
46.45
42.83
39.43
47.38
42.81
39.6
47.12
42.86
38.92
47.66
42.86
38.8
Liner5
Perc.
4.31
4.22
4.12
4.33
4.22
4.1
3.12
3.08
3.04
3.12
3.08
3.0
Runoff
8.57
7.87
7.06
10.48
7.87
5.22
4.36
3.45
2.98
6.63
3.45
2.32
ET3
49.78
53.55
57.18
50.40
53.55
57.34
54.57
57.05
59.99
55.24
57.05
59.60
Liner6
Perc.
42.16
39.41
37.02
40.02
39.41
38.20
40.08
38.84
36.69
37.65
38.84
37.49
1CA = Santa Maria, CA; LA = Shreveport, LA; DP = drainable porosity (vol/vol); PAWC = plant available water capacity
(vol/vol). All cases are for 46 cm (18 in.) of sandy loam topsoil (HELP model default texture 6); fair grass; evaporative depth
= 25 cm (10 in.); and curve number = 75.
2Change in storage is not included in this table; therefore, the water balance components shown do not always add up to
100.0 percent.
n
ET = evapotranspiration.
4Lateral drainage from a 31-cm (12-in.) layer having aslope of 3 percent, a drainage length of 61 m (200ft), and a hydraulic
conductivity of 3 x 10~2 cm/sec.
5Percolation through 61-cm (24-in.) liner having a hydraulic conductivity of 10 cm/sec.
6Percolation through 61-cm (24-in.) liner having a hydraulic conductivity of 10~6 cm/sec.
bulk volume of soil including voids. The values for wilting
point and drainable porosity remained constant at 0.338
and 0.457 cm/cm (0.133 and 0.180 in./in.), respectively.
Only sandy loam soil was considered. The evaporative
depth was 25 cm (10 in.), and the SCS runoff curve num-
ber was 75. Both two- and three-layer cover designs
were simulated. The results are presented in Table 9-7.
Increasing the plant available water capacity provides a
greater volume of water available for evapotranspiration
after vertical drainage has nearly ceased. This results in
larger volumes of evapotranspiration as shown in Table
9-7. Consequently, the lateral drainage and percolation
decreases. The change in the volume of runoff was
design dependent. Since increasing the plant available
water capacity results in an increased moisture content
at field capacity, there is a greater potential for higher an-
tecedent moisture conditions or relative moisture content,
resulting in a higher curve number. As such, the runoff
for the three-layer cover systems increased with increas-
ing plant available water capacity. Runoff decreased for
the two-layer cover systems because infiltration is limited
by the storage volume above the liner. As such, increas-
ing the plant available water capacity increases the
storage volume, reducing the limits on infiltration and the
runoff. As shown in Table 9-7, the runoff from the two-
layer cover approaches the runoff from the three-layer
cover as the storage potential in the two-layer cover
becomes large, that is for large values of drainable
porosity and plant available water capacity. In all
cases the increases in evapotranspiration were
great enough to offset any decrease in runoff; therefore,
leachate drainage and percolation always decreased.
The size of the changes in the water budget components
were dependent on the climate and design. The results
would also be dependent on the type of topsoil.
Liner/Drain Systems
This section examines the effects of liner/drain system
design on the performance of the drain system under
conditions typical of cover systems, and leachate collec-
tion systems in open and closed landfills. Performance
was determined by the apportionment of the drainage
into the drain layer between lateral drainage and percola-
tion through the liner. In addition, the effect of design on
the resulting depth of saturation also was examined. For
106
-------
Table 9-8. Sensitivity of Lateral Drainage and Liner Percolation to Lateral Drainage Slope and Length
Annual1
Infill.
(in.)
50
50
50
50
50
50
50
50
8
8
8
8
8
8
8
8
Avg. Annual Vol.
(% Inflow)
Slope
S
(ft/ft)
0.01
0.01
0.01
0.03
0.03
0.03
0.09
0.09
0.01
0.01
0.01
0.03
0.03
0.03
0.09
0.09
Length
L
(ft)
25
75
225
25
75
225
25
75
25
75
225
25
75
225
25
75
S*L
(ft)
0.25
0.75
2.25
0.75
2.25
6.75
2.25
6.75
0.25
0.75
2.25
0.75
2.25
6.75
2.25
6.75
US
(ft)
2,500
7,500
22,500
830
2,500
7,500
280
830
2,500
7,500
22,500
830
2,500
7,500
280
830
Lat.2
Drng.
96.71
95.89
93.43
96.85
96.36
95.10
97.37
96.87
83.73
82.29
78.51
84.16
83.59
82.28
84.35
84.23
Liner3
Perc.
3.29
4.11
6.57
3.15
3.64
4.90
2.63
3.13
16.27
17.71
21.49
15.84
16.41
17.72
15.65
15.77
Max. Head
In Lat.
Drng. Layer
(in.)
13.8
29.7
58.2
12.3
24.8
42.3
8.5
16.2
1.2
3.4
9.4
0.5
1.1
3.5
0.2
0.4
1Value of 50 in./yr represents inflow through an open landfill; the temporal distribution is based on rainfall records for Shreveport,
LA. Value of 8 in./yr represents inflow through landfill cover; the temporal distribution is uniform throughout the year.
2Lateral drainage from a layer having a slope of 3 percent, drainage length of 75 ft, porosity of 0.351 vol/vol, field capacity of
0.174 vol/vol, and a saturated hydraulic conductivity of 10~2 cm/sec
3Percolation through a 24-in.-thick soil liner having a saturated hydraulic conductivity of 10"7 cm/sec.
the cover system or open landfill the drainage into the
drain layer was 127 cm/yr (50 in./yr), distributed tem-
porally in accordance with the precipitation at
Shreveport. For the closed landfill the drainage into the
drain layer was distributed uniformly through time at a
rate of 20 cm/yr (8 in./yr).
Four types of liner/drain systems are examined in the
various parts of this study to determine their perfor-
mance: a sand drainage layer underlain by a clay liner, a
sand drainage layer underlain by a geomembrane, a
sand drainage layer underlain by a composite liner, and
double liner systems. For the clay liner system this sen-
sitivity analysis determines the effects of the saturated
hydraulic conductivity of the liner and drain layer, slope of
the liner, and drain spacing. For the geomembrane and
composite liner systems, the effects of synthetic liner
leakage fraction and saturated hydraulic conductivity of
the geomembrane's subsoil are examined. The sen-
sitivity of the parameters affecting the synthetic liner
leakage fraction are presented graphically. For the
double liner systems, the effectiveness of several dif-
ferent systems in preventing and detecting leakage from
the primary liner prior to reaking through the secondary
liner was compared. In all systems the thickness of the
drain layer was greater than the peak depth of saturation
in the drain layer, and the thickness of the clay liner or
subsoil below a geomembrane was 61 cm (24 in.).
Clay Liner/Drain Systems
Saturated Hydraulic Conductivities. The liner/drain sys-
tem used in this analysis is shown as Design A in Figure
9-10. The value of KD (the saturated hydraulic conduc-
tivity of the drain layer) ranged from 0.001 to 1 cm/sec
while the value of KP (the saturated hydraulic conduc-
tivity of the clay liner) ranged from 10 to 10~5 cm/sec.
The slope of the liner surface toward the drainage collec-
tor was 3 percent, and the maximum drainage length to
the collector was 23 m (75 ft). The results of the
drainage efficiency determinations for the various com-
binations of KD and KP are shown in Figure 9-4, where
the average annual volumes of lateral drainage and per-
colation expressed as a percentage of annual inflow are
plotted.
For the large unsteady inflows totaling 127 cm/yr (50
in./yr), only designs where the saturated hydraulic con-
ductivity of the liner was equal to or less than 10"7 cm/sec
limited the percolation through the liner to volumes less
than 5 percent of the annual inflow (6.4 cm [2.5 in.]). The
effect of KD on the drainage efficiency for these low per-
meability liners is fairly small. Changing KD from 0.001
cm/sec to 1 cm/sec reduced the percolation from 7 per-
107
-------
cent to 1 percent of the inflow for a KP of 10~7 cm/sec
and from 0.7 percent to 0.1 percent for a KP of 10~8
cm/sec. For a KP value of 10 cm/sec, only a KD value
of 1 cm/sec or greater can reduce the percolation to less
than 10 percent of the annual inflow. Liners having a KP
of 10"5 cm/sec are largely ineffective no matter how large
the value of KD is.
For smaller steady inflows of 20 cm/yr (8 in./yr) typical
of the infiltration through some cover systems, only
liners having a value of KP equal to or less than 10
cm/sec limited leakage except for designs having a KP
of 10~6 cm/sec and a very large KD value, 1 cm/sec or
greater. As above, the effect of KD on the drainage ef-
ficiency is small. Changing KD from 0.001 cm/sec to 1
cm/sec reduced the percolation from 22 percent to 15
percent of the inflow for a KP of 10~7 cm/sec and from
2.3 percent to 1.5 percent for a KP of 10~8 cm/sec.
Liners having a KP of 10~7 cm/sec leaked between 2.5
and 5.1 cm/yr (1 and 2 in./yr) while liners having a KP
of 10~8 cm/sec leaked between 0.25 and 0.51 cm/yr
(0.1 to 0.2 in./yr).
Summarizing the results shown in Figure 9-4, the
saturated hydraulic conductivity of the liner is the primary
control of leakage through a clay liner. At hydraulic con-
ductivities below about 10"6 cm/sec the leakage is nearly
proportional to the value of KP; that is, an order of mag-
nitude decrease in the value of KP yields nearly an order
of magnitude decrease in percolation. The value of KD
has only a small effect on the leakage through liners
having a KP of 10"7 cm/sec or less. Changing the value
of KD by three orders of magnitude when using these low
permeability liners yields much less than an order of
magnitude change in percolation.
Similar effects are also seen in Figures 9-5 and 9-6
which relate the KD/KP design ratio to the resulting ratio
of lateral drainage to percolation. The curves in Figure 9-
5 are log-least-squares regressions for several ranges of
steady-state heads resulting from a steady-state inflow of
20 cm/yr (8 in./yr). The curves in Figure 9-6 are log-least-
squares regressions for several ranges of peak "y result-
ing from a unsteady inflow of 127 cm/yr (50 in./yr). The
plotted points are QD/QP ratios for the given KD/KP
ratio; their symbols indicate the value of KD used in ob-
taining the result. The actual steady-state "y and peak "y
values were both grouped into four ranges of heads. In
Figure 9-5 steady-state heads ranging from 26 to 30.7
cm (10.2 to 12.1 in.) were grouped together as were
heads ranging from 3.56 to 4.06 cm (1.4 to 1.6 in.),
equaling 0.508 cm (0.2 in.), and less than 0.127 cm (0.05
in.). In Figure 9-6 peak heads ranging from 6.1 to 6.4 cm
(2.4 to 2.5 in.) were grouped together as were heads
ranging from 19.3 to 23.6 cm (7.6 to 9.3 in.), from 41.15
to 69.6 cm (16.2 to 27.4 in.), and from 116.1 to 153.2cm
(45.7 to 60.3 in.).
Figures 9-5 and 9-6 show that percolation tends to
dominate at ratios of KD/KP below 107. This is par-
ticularly true as the depth of saturation or inflow
decreases. When heads remain constant, the ratio of
lateral drainage to percolation is a linear function of
KD/KP. Using the maximum head allowed by RCRA of
31 cm (12 in.) and the current minimum KD/KP ratio im-
plied by RCRA of 105, a percolation of 2.3 percent of in-
flow results; however, an unusually large steady-state
inflow of 203 cm/yr (80 in./yr) or 0.559 cm/day (0.22
inVday) is required to achieve this condition. When using
the RCRA guidance design, therefore, the peak and
steady-state average heads will be considerably smaller
than 31 cm (12 in.) at virtually all locations.
Slope and Drainage Length. The combinations of slope
and drainage length used in this analysis are listed in
Table 9-8 along with resulting average annual volumes of
lateral drainage and percolation expressed as a percent-
age of annual inflow. The table also contains the result-
ing maximum heads above the soil liner. The slope (S)
ranged from 0.003 to 0.028 cm/cm (0.01 to 0.09 ft/ft) (1 to
9 percent) while the drainage length (L) ranged from 8 to
69 m (25 to 225 ft). The saturated hydraulic conduc-
tivities of the lateral drainage and soil liners were 10"2
and 10"7 cm/sec, respectively. The product S*L and the
ratio US ranged from 0.76 to 2 m (0.25 to 6 ft) and 85 to
6,858 m (280 to 22,500 ft), respectively. S*L is the head
contributed by the liner at the crest of the drainage layer.
The results indicate that the volumes of lateral drainage
and percolation vary little with changes in slope and
drainage length under both steady and unsteady inflows.
A ninefold increase in slope reduced the percolation by a
maximum of 25 percent for the unsteady inflow and 13
percent for the steady inflow. As the drainage length is
reduced and the slope increased, the lateral drainage
rate increases. As a result, the head decreases and is
maintained at smaller depths for shorter durations. Con-
sequently, the percolation decreases since it is a function
of the head on the liner. A ninefold decrease in drainage
length reduced the percolation by a maximum of 50 per-
cent for the unsteady inflow and 25 percent for the steady
inflow. A ninefold increase in slope and decrease in
length decreased the percolation by about 60 percent for
the unsteady inflow and about 30 percent for the steady
inflow.
The head in the drain layer varies greatly with changes in
slope and drainage length. For a steady inflow the
average head increases linearly with an increase in
drainage length and an increase in the inverse of the
slope, as shown in Figure 9-7. A similar relationship ex-
ists between the peak average head during the simula-
tion and L/S for unsteady inflow. The average head is
slightly influenced by the product of the slope and
drainage length when the head is similar to this product.
108
-------
KD = I cm/s
L KD = 0.1 cm/s
KD = 0.0! cm/s \\\
KD = 0.001 cm/s
o 50 in./yr Inflow
8 In./yr SS Inflow
10 ° 10'' 10 u 10
KP Ccm/s)
Figure 9-4. Effect of saturated hydraulic conductivity on lateral drainage and percolation.
Geomembrane/Drain Systems
A single synthetic liner under a drain layer as shown in
Design B in Figure 9-10 is examined in this section. It is
assumed that the synthetic liner was laid directly on a
3-m (10-ft) thick layer of native subsoil. The drainage
layer had a saturated hydraulic conductivity of 10~2
cm/sec, a slope of 3 percent, and a drainage length of
23 m (75 ft). This case will be used to demonstrate the
influence of the synthetic liner leakage fraction and the
saturated hydraulic conductivity of the native subsoil on
the liner system performance. The properties of the sub-
soil ranged from sand to clay in the analysis.
Liner Leakage Fraction. Brown et al. (4) conducted
laboratory experiments and developed predictive equa-
tions to quantify leakage rates through various size holes
in synthetic liners over soil. They assumed that the
measured leakage rates corresponded to a uniform ver-
tical percolation rate equal to the saturated hydraulic con-
ductivity through a circular cross-sectional area of the soil
liner directly beneath the hole. Using the data relating
leakage and cross-sectional area of flow, Brown et al. (4)
developed predictive equations for the radius or area of
this flow cross section as a function of hole size, depth of
leachate ponding, and saturated hydraulic conductivity of
the soil. Figure 9-8 presents their results. The radius of
saturated flow through the subsoil was significantly
greater than the radius of the hole in the synthetic liner.
In this paper, the cross-sectional area of saturated flow
was multiplied by the number of holes per unit area of
synthetic liner to compute the synthetic liner leakage frac-
tion. Liner leakage fraction is simply defined as the total
horizontal area of saturated flow through the subsoil
beneath all of the liner holes divided by the horizontal
area of the liner.
Liner leakage fraction is a function of many parameters,
some quantitatively defined and others qualitatively
defined. Liner leakage fraction increases linearly with in-
creases in the number of holes of the same size and
shape. Shape also has a strong effect on the leakage;
tears have larger leakage than punctures. Increasing the
size of circular holes yields only a slight increase in the
leakage, while increasing the length of a tear or bad
seam increases the leakage nearly linearly. Leakage
also increases nearly linearly with increases in head or
depth of saturation above the liner. The leakage fraction
also is affected by the gap width between the liner and
the subsoil. Gap width is a measure of the seal between
the liner and the subsoil. The smaller the gap the better
the seal. The seal is a function of the subsoil, installa-
tion, liner placement, and subsoil preparation. Installa-
tion of the liner on coarse-grained subsoil, clods, debris,
or filter fabric provides a poor seal as will wrinkles in the
liner. Coarse-grained subsoils decrease the leakage
fraction while greatly increasing the leakage. The greater
permeability of coarse materials allows greater flow
through a smaller area of saturated flow, reducing the
109
-------
Q.
a
a
a 10
8 In./yr SS Inflow
* KD - O.OO! cm/s
O KD - O.OI cm/s
n KD - O.1 cm/Q
O KD - 1 cm/s .#
SS y < 0.1 In.
SS y - 0.2 In.
S3 y = 1.5 In.
SS u =s 1 I In.
10
10 10 1O
KD/KP
8
1O
1O
Figure 9-5. Effect of ratio of drainage-layer saturated hydraulic conductivity to soil-liner saturated hydraulic conductivity
on ratio of lateral drainage to percolation for steady-state (SS) inflow of 20 cm/yr (8 in./yr).
Q.
Q
\
a
a 10
BO In./yr Inflow
/ o KD - I cm/s
a KD - O.I cm/s
O KD = O.OI cm/s
* KD = O.OOI cm/s
Peak y = 2.5 In.
Peak y = Q In.
Peak y = 24 In.
Peak y = 55 In.
10
10 1O 1O
KD/KP
8
IO
10
8
Figure 9-6. Effect of ratio of drainage-layer saturated hydraulic conductivity to soil-liner saturated hydraulic conductivity
on ratio of lateral drainage to percolation for unsteady inflow of 127 cm/yr (50 in./yr).
110
-------
Q
-------
o
a:
CD
a.
CD
c
c
CD
a.
a
CD
XI
10'
10-
10'
10°
10-
Uppsr bound is for 0.08-cm-dia. openings.
Lower bound is for I.27-cm-dia. openings.
2H KP = 3.4 x 10
10
,-6
10-5
10'
,-4
10
-3
10
-2
10-
10
20
50
100
200
500
10°
0)
CD
c
Q.
O
C
O
CD
3.
-t-»
0)
CO
en
C
o
a
a.
en
E
i_
O
Synthetic Liner Leakage Fraction. LF
Figure 9-8. Synthetic liner leakage fraction as a function of density of holes, size of holes, head on the liner, and saturated
hydraulic conductivity of the liner.
100
O * / ^ ' *- /
0:/ : / • n. • /
to
,0-10 ,o-9 ,Q-8
LF x KP (cra/s)
Figure 9-9. Effect of leakage fraction on system performance.
112
-------
layer and vertical percolation through each synthetic liner
and each soil liner. These predictions were based on 20
cm/yr (8 in./yr) of infiltration passing through the waste
layer and reaching the primary leachate collection sys-
tem. This inflow was distributed uniformly in time.
Figures 9-11 and 9-12 show the results in terms of lateral
drainage from the secondary drainage layer and vertical
percolation through the bottom soil liner as functions of
synthetic liner leakage fraction of the top membrane.
Design C consists of a primary leachate drainage layer
underlain by a synthetic liner, a secondary drainage
layer, and a soil liner. As shown in Figure 11, this design
is not very effective. Large quantities of leakage oc-
curred at fairly low leakage fractions and no leakage
(lateral drainage) was detected from the secondary
drainage layer until the synthetic liner leakage fraction
exceeded about 10~5. At smaller synthetic liner leakage
fractions, the leachate percolated vertically through the
soil liner as fast as the leakage through the synthetic liner
occurred. The product of the saturated hydraulic conduc-
tivity of the secondary drainage layer times the synthetic
liner leakage fraction must be greater than or ap-
proximately equal to the saturated hydraulic conductivity
of the soil liner before leakage will be detected using this
design. At the time leakage is detected, the vertical per-
colation rate through the soil liner could be about 16 per-
cent of total inflow.
Design D consists of a primary drainage layer underlain
by a synthetic liner, a soil liner, a secondary drainage
layer, and a second soil liner. The soil liner immediately
below the synthetic liner is very effective in minimizing
vertical percolation (leakage through the primary liner);
however, a synthetic liner leakage fraction greater than
10"2 to 10~1 would be required before leachate would be
collected from the secondary drainage layer. Because
the vertical percolation through the first liner is so small,
practically all of the leakage is removed by vertical per-
colation through the bottom soil liner as shown in Figure
9-12. This design is ineffective since the leakage detec-
tion system would not function.
Design E consists of a primary drainage layer underlain
by a synthetic liner, a secondary drainage layer, a
second synthetic liner, and a soil liner. In this case, any
leakage through the upper synthetic liner will readily pass
through the underlying drainage medium to the lower
synthetic liner. Since the lower synthetic liner is under-
lain by a soil liner, most leakage will be collected by
lateral drainage. Figure 9-11 shows that leakage will be
detected far in advance of significant vertical percolation
from the landfill. That is, the leakage fraction of the syn-
thetic liners at which leakage detection will occur is
several orders of magnitude smaller than the leakage
fraction at which significant vertical percolation from the
landfill will occur. The leakage lost by percolation is vir-
DESIGN A
DESIGN B
DESIGN C
',:^WASTE\AYER:r
DRAIN LAYER
\\ \ \ v
v \SOIL LINER \
\ \ \ \ V
DRAIN LAYER
—SYNTHETIC L INER
\ \ \ \ \
\NATIVE SUBSOIL
\ \ \
^
"•,WASTE""LAYER?,J(
;DRAIN LAYER
_( ^^—SYNTHETIC LINER
DRAIN LAYER
\\ \ \ \
\ SOIL LINER \
\ \ \ \ \
DESIGN D
DESIGN E
DESIGN F
#'.VASTE LAYER °
Q.-^yo- °y :c.'~<:
DRAIN LAYER
'^SYNTHETIC LINER
\ \ \ \ \^
v NSOIL LINER \
\ \ \ \ \X
DRAIN LAYER
\ \ \ \\%
XSOIL LINER \
\\v\\.
.
n-WASTE LAYER/
••°''-''-'V°<£3
DRAIN LAYER
^-SYNTHETIC. LINER
DRAIN LAYER.
^-SYNTHETIC LINER
\ \ \ \
^SOIL LINER N
. \ \ \ V
,,*-• - ,*.c
f.lWASTE LAYER *
,Gi UOO.."Q. --^-^? " "A?
•DRAIN LAYER
C L INER
"\\ \ \ V
SOIL LINER \
\ \ \ \ \
::;DRAIN LAYERS.
^l+— SYNTHETIC LINER
\ \ \ \
-SOIL LINER \
\ \ \ \
Figure 9-10. Liner designs.
113
-------
inn
80
r-»
^
o
- 60
c
^ 40
k_/
3
20
0
(_ __ — Design E
/ ^-.
QQ t. Design C
i /
1 1
1 1
1 1
! I
'
I / Design C
__^// __X
Design E "
10
-7
10"° 10'° 10"" 10"° 10"^ 10"' IOL
Synthetic Liner Leakage Fraction, LF
Figure 9-11. Percent of inflow to primary leachate collection layer discharging from leakage detection layer and bottom
liner double-liner systems C and E.
o
»•—
C
O.OI O.IO I.00
Synthetic Liner Leakage Fraction. LF
Figure 9-12. Percent of inflow to primary leachate collection layer discharging from leakage detection layer and bottom
liner for double-liner systems D and F.
114
-------
tually the same as for Design D but detection is much
better. This design is effective at minimizing leakage
from the landfill and at detecting leakage through the
primary liner, but significant leakage through the primary
liner may occur at fairly low liner leakage fractions.
Design F consists of a primary drainage layer underlain
by a synthetic liner, a soil liner, a secondary drainage
layer, a second synthetic liner, and a second soil liner.
Figure 9-12 shows that the addition of the lower synthetic
liner improves the system performance in comparison to
the performance of Design D. Leakage is detected
whenever leakage occurs. Even at leakage fractions of
10~3 when only 0.02 percent of the inflow leaks through
the primary liner, half of the leakage is collected in the
secondary drainage layer. The depth of saturation in the
secondary drainage layer is lower than in the primary
layer. This sufficiently reduces the leakage through the
second synthetic liner to permit detection whenever the
primary liner leaks. Design F is a very effective double-
liner design because it minimizes the leakage through
the primary liner and from the landfill and collects
leakage at all leakage fractions.
A comparison of the four designs shows that Design F is
the most effective in detecting the earliest leaks with the
least amount of vertical leakage through the primary liner
and also through the bottom soil liner. Design D yields
the same quantity of leakage through the primary liner;
however, leakage in Design D would probably never be
detected or collected. Therefore, the bottom liner in
Design D is not functional. Designs D and E yield the
same leakage through the bottom liner but Design E
detects leakage through the primary liner at the lowest
leakage fraction. Design C also detects leaks at very
small leakage fractions but allows significant vertical per-
colation through the bottom soil liner before detection.
The leakage through the primary liner in Designs C and
E is large even at low leakage fractions. Therefore, syn-
thetic membranes placed on highly permeable subsoils
are ineffective except for very low inflows and for very
low leakage fractions. Synthetic membranes are best
used in conjunction with a low-permeability soil as a com-
posite liner. Comparison of the results for Designs B and
C demonstrates this point. Both designs are composed
of one synthetic membrane and one soil liner, but the
leakage from the composite liner (Design B) shown in
Figure 9-9 as the curve for 20 cm/yr (8 in./yr) steady in-
flow is much lower than the leakage from the double liner
system (Design C) as shown in Figure 9-11.
It is interesting to compare the single-liner performance
of Design B to the double-liner performance of Design D,
assuming the soil-liner-saturated hydraulic conductivity in
Design B is the same as Design D. The vertical percola-
tion leaving the system in Design B is essentially the
same as that leaving the secondary liner in Design D as
seen by comparing Figure 9-12 to the curve in Figure 9-9
for 20 cm/yr (8 in./yr) steady inflow. The secondary liner
in Design D is nonfunctional since the percolation rate of
the second soil liner is generally equal to or greater than
the leakage rate.
SUMMARY OF SENSITIVITY ANALYSIS
The interrelationship between variables influencing the
hydrologic performance of a landfill cover is complex. It
is difficult to isolate one parameter and exactly predict its
effect on the water balance without first placing restric-
tions (sometimes severe restrictions) on the values of the
remaining parameters. With this qualification in mind, the
following general summary statements are made.
The primary importance of the topsoil depth is to control
the extent or existence of overlap between the evapora-
tive depth and the head in the lateral drainage layer. The
greater this overlap, the greater will be evapotranspira-
tion and runoff. Surface vegetation has a significant ef-
fect on evapotranspiration from soils with long
flow-through travel times and large plant available water
capacities; otherwise, the effect of vegetation on
evapotranspiration is small. The general influence of sur-
face vegetation on lateral drainage and percolation is dif-
ficult to predict outside the context of an individual cover
design. Clay soils increase runoff and evapotranspiration
and decrease lateral drainage and percolation. Simula-
tions of landfills in colder climates and in areas of lower
solar radiation are likely to show less evapotranspiration
and greater lateral drainage and percolation. An in-
crease in the runoff curve number will increase runoff and
decrease evapotranspiration, lateral drainage, and per-
colation. As evaporative depth, drainable porosity, or
plant available water increase, evapotranspiration tends
to increase and lateral drainage and percolation tend to
decrease; the effect on runoff is varied.
The sensitivity analysis shows that the ratio of lateral
drainage to percolation is a positive function of the ratio
of KD/KP and the average head above the liner.
However, the average head is a function of QD/QD and
US. The quantity of lateral drainage, and, therefore, also
the average head, is in turn a function of the infiltration.
Therefore, the ratio of lateral drainage to percolation in-
creases with increases in infiltration and the ratio of
KD/KP for a given drain and liner design. The ration of
lateral drainage to percolation for a given ratio of KD/DP
increases with increases in infiltration and the term S/L.
The percolation and average head above the liner is a
positive function of the term US.
Leakage through geomembrane increases with the num-
ber and size of holes, the depth of water buildup on the
liner, the permeability of the subsoil, and the gap bet-
ween the liner and the subsoil. Geomembranes reduce
leakage through liner systems by reducing the area of
saturated flow through the subsoil. The overall effective-
ness of a geomembrane system is equivalent to a soil
liner having a saturated hydraulic conductivity equal to
the product of the saturated hydraulic conductivity of the
115
-------
subsoil and the ratio of the reduced area of flow through
the subsoil to the area of the liner. Composite liners
provide the best reduction in leakage. Drain systems
that yield low head buildup on the geomembrane improve
the performance of a geomembrane system.
REFERENCES
1. Schroeder, P. R., and Peyton, R. L. 1987. "Verifica-
tion of the Hydrologic Evaluation of Landfill Perfor-
mance (HELP) Model Using Field Data." EPA 600/2-
87-050. EPA Hazardous Waste Engineering Re-
search Laboratory, Cincinnati, OH.
2. Schroeder, P. R., R. L. Peyton, and J. M. Sjostrom.
1988. Hydrologic Evaluation of Landfill Performance
(HELP) Model: Vol. Ill User's Guide for Version 2.
Internal Working Document. USAE Waterways Ex-
periment Station, Vicksburg, MS.
3. U.S. Department of Agriculture, Soil Conservation
Service. 1972. Section 4, Hydrology. In: National
Engineering Handbook, U.S. Government Printing
Office, Washington, DC. 631 pp.
4. Brown, K.W., J.C. Thomas, R.L. Lytton, P. Jayawik-
rama, and S.C. Bahrt. 1987. Quantification of Leak
Rates Through Holes in Landfill Liners. EPA/600/S2-
87-062. EPA Office of Research and Development,
Cincinnati, OH.
5. Peyton, R. L. and Schroeder, P. R. 1990. "Evalua-
tion of Landfill-Liner Designs." Vol. 116, No. 3, Jour-
nal of Environmental Engineering Division, American
Society of Civil Engineers.
116
-------
CHAPTER 10
GAS MANAGEMENT SYSTEMS
GAS GENERATION
The information in this chapter applies mainly to Subtitle
D landfills. Hazardous waste landfills (Subtitle C) do not
usually contain significant amounts of organic materials
and, thus, normally have a minimal gas management
system as a component of the final cover.
Gas generation in a landfill system poses several
problems. If allowed to accumulate, gas is an explosion
hazard. It also provides stress to vegetation by lowering
the oxygen content available at the roots, severely affect-
ing the ability of the cover to support vegetation. In the
absence of adequate corridors for the gas to escape, gas
pressures can increase sufficiently to physically disrupt
the cover system as well, generating large cracks and
rupturing the geomembrane. Other problems include
odor, toxic vapors, and uncontrolled gas migration which
can cause deterioration of nearby property values.
Gas generation is a product of anaerobic decomposition
of organic materials placed in the landfill. The decom-
position can be described by the reaction:
wCO2
+ humus (1)
The composition of landfill gas generally is about 50 per-
cent methane, 40 percent carbon dioxide, and 10 percent
other gases including nitrogen products. This particular
mix of gases generally will not occur until after the landfill
becomes anaerobic. During the first year after the
materials are placed in the landfill, the gas is
predominantly carbon dioxide and is unsuitable for
recovery and use. After the methane content rises, the
gas can be mined as a fuel or energy source. However,
the BTU value of landfill gas is about half that of natural
gas and, therefore, is generally too low to substitute
directly for natural gas. Landfill gas requires purification
and is frequently used in conjunction with natural gas.
Waste decomposition rates and hence gas production
rates are moisture dependent. Highest gas production
rates occur at moisture contents ranging from 60 to 80
percent of saturation. In modern landfill design, infiltration
of water into the waste is restricted to a practical mini-
mum; therefore, optimum moisture contents may never
be achieved. Consequently, gas production rates may be
much lower than anticipated, decreasing the attractive-
ness of gas recovery systems. To maximize gas produc-
tion, strategies such as leachate recirculation should be
employed to distribute bacteria, nutrients, and moisture
more uniformly. Typical gas production rates from wet,
anaerobic wastes are about 20 to 50 mL/kg/day. These
high production rates will continue for decades. Produc-
tion at low rates may continue for centuries because of
quantity of material and resistance of some material to
biodegradation.
GAS MIGRATION
Gas migrates from landfills through two mechanisms—
convection and diffusion. Convection is transport induced
by pressure gradients formed by gas production in layers
surrounded by low hydraulic conductivity or saturated
layers. Convection also results from buoyancy forces be-
cause methane is lighter than carbon dioxide and air.
Diffusion is the transport of materials induced by con-
centration gradients. Anaerobic decomposition produces
a gas mixture with concentrations of methane and carbon
dioxide that are much greater than those found in the sur-
rounding air. Therefore, molecules of methane and car-
bon dioxide will diffuse from the landfill gas to the air in
accordance with Pick's law. Diffusion plays a much
smaller role in gas migration than convection.
Many factors affect gas migration. Some of the more im-
portant factors are the landfill design, including refuse cell
construction; final cover design; and incorporation of gas
migration control measures. Low hydraulic conductivity
soil layers and geomembranes are very effective barriers
to gas migration. Sand and gravel layers and void spaces
provide effective corridors for channeling gas migration.
Other channels affecting migration are cracks and fis-
sures between and in lifts of waste or soil due to differen-
tial settlement and subsidence.
Other factors affecting gas migration include the gas
production rate, the presence of natural and artificial con-
duits and barriers adjacent to the landfill, and climatic and
seasonal variations in site conditions. High gas produc-
tion rates increase migration. Corridors at the site ad-
jacent to a landfill such as water conduits, drain culverts,
buried lines, and sand and gravel lenses, promote uncon-
117
-------
trolled migration from the site. Barriers can include clay
deposits; high or perched water tables; roads; and com-
pacted, low hydraulic conductivity soils. Environmental
variations can result from the intermittent occurrence of
saturated or frozen surface soils, which seals the surface
and promotes lateral migration. Barometric pressure
changes also affect the rate of gas release to the surface.
Seasonal changes in moisture content can change the
gas production rate and, therefore, the extent and quan-
tity of migration.
GAS CONTROL STRATEGIES
Two gas control strategies—passive and active—are
available, and may be used at any facility. Passive sys-
tems provide corridors to intercept lateral gas migration
and channel the gas to a collection point or a vent. These
systems use barriers to prevent migration past the inter-
ceptors and the perimeter of the landfill. Active systems
generate a zone of negative pressure to increase the
pressure gradient and, consequently, the flow toward the
zone. Active systems also can be used to create a zone
of high pressure to prevent gas migration toward the
zone.
Typical passive systems are shown in Figures 10-1,10-2,
10-3, and 10-4. Figure 10-1 shows a gas-vent layer used
in conjunction with a composite liner and vent in the
cover system (2). The composite liner prevents uncon-
trolled vertical migration, while the gas-vent layer inter-
cepts all vertical migration and directs it to the vent.
Figure 10-2 shows a gravel vent that runs diagonally
down through the waste material. The gravel vent inter-
cepts both vertical and lateral migration and channels it
to the surface. Figures 10-2,10-3, and 10-4 show gravel-
filled trenches (1, 3, 4). The trenches intercept lateral
migration and direct the gas to the surface where it is
vented or extracted. Gravel-filled trenches on the
perimeter of the landfill are often used with an imperme-
able barrier on the outer side of the trench to prevent
migration from the trench to the surrounding area. These
systems often extend from the surface down to a low
hydraulic conductivity soil layer or other barrier such as
the water table or a geomembrane. The systems may be
as deep as the bottom of the landfill, or even lower if out-
side the landfill. Extreme care should be taken in the
design of all of these vent systems to prevent them from
being a source of infiltration through the cover. Improper
design could allow the vent to intercept surface runoff
and pipe additional infiltration into the leachate collection
system.
Typical active systems are shown in Figures 10-4, 10-5,
and 10-6 (4). All three figures show gas extraction wells
using exhaust blowers. The well is placed in a gravel vent
or gravel-filled trench located in the waste cell (Figures
10-5 and 10-6) or along its perimeter (Figure 10-4). The
gravel vent is sealed to prevent the well from drawing air
from the surface and destroying the suction (zone of
negative pressure) needed to draw gas to the well. The
seal also prevents infiltration of surface water. Imperme-
able barriers in the cover and perimeter walls increase
the efficiency of gas extraction wells since they restrict in-
flow of air that would dissipate the suction. In addition, it
reduces the number of wells needed and increases the
heating value of the gas collected. Typically, gas extrac-
tion wells do not extend to the bottom of the landfill since
the suction is able to draw gas from a sizable zone
beyond the gravel fill.
gas vent
drain layer
geomembrane
vent layer
JuM^M
top layer
low-permeability
geomembrane/soil layer
waste
Figure 10-1. Cover with gas vent outlet and vent layer.
118
-------
~— SLOPE
FINAL COVER MATERIAL
Figure 10-2. Gravel vent and gravel-filled trench used to control lateral gas movement in a sanitary landfill (5).
WINTER CLIMATES MAY REQUIRE
COLLECTOR WITH VERTICAL RISERS
AND SURFACE SEAL
GRAVEL BACKFILL'
BARRIER
MATERIAL
(IMPERVIOUS
MEMBRANE)
COVER MATERIAL
REFUSE
UNDISTURBED IMPERVIOUS MATERIAL OR WATER TABLE "
Figure 10-3. Typical trench barrier system.
119
-------
"A"
Permeable Trench
\ Perforated
Pipe
"C"
Pipe Vent
LEGEND
Gas Migration
Refuse
Gravel
Trench Cover
Impermeable Layer
"B"
Impermeable Barrier
Exhaust
Blower
Perforated
Pipe
"D"
Induced Exhaust
Gas Control Barriers
Figure 10-4. Gas control barriers (6).
120
-------
GAS FLOW
GAS FLARE
EXHAUST BLOWER
IMPERVIOUS BACKFILL
^-PERFORATED PIPE
GAS FLOW
PERMEABLE MATERIAL
Figure 10-5. Gas extraction well for landfill gas control.
REFERENCES
1. Lutton, R.J., G.L. Regan, and L.W. Jones. 1979.
Design and construction of covers for solid waste
landfills. EPA-600/2-79-165 U.S. EPA Municipal En-
vironmental Research Laboratory, Cincinnati, OH.
2. U.S. EPA. 1989. Technical guidance document: final
covers on hazardous waste landfills and surface im-
poundments. EPA/530-SW-89-047.
3. Shafer, R.A., A. Renta-Babb, J. T. Bandy, E. D.
Smith, and P. Malone. 1984. Landfill gas control at
military installations. Technical Report N-173. U.S.
Army Engineer Construction Engineering Research
Laboratory, Champaign, IL.
4. McAneny, C.C., P.G. Tucker, J.M. Morgan, C.R. Lee,
M.F. Kelley, and R.C. Horz. 1985. Covers for uncon-
trolled hazardous waste sites. EPA/540/2-85/002
U.S. EPA Hazardous Waste Engineering Research
Laboratory, Cincinnati, OH.
5. Brunner, D.R. and D.J. Keller. 1971. Sanitary landfill
design and operation. SW-66TS. U.S. Environmental
Protection Agency.
6. Rovers, F.A., J.J. Tremblay, and H. Mooij. 1977. Pro-
cedures for landfill gas monitoring and control. EPS
4-EC-77-4, Waste Management Branch, Environ-
ment Canada, Ottawa.
121
-------
COLLECTION HEAOCR
•^^H^ TELESCOPIC COUPLING
•v;-'.7.'v's'cH'eo pvc. V-6"0
' i'/'<• PERFORATED PIPE
-IECEND-
REFUSE
FINAL COVER - Z'-e'
CLAY PLUO
FINE SAND
COARSE GRAVEL
Gas Extraction
Well Design
Figure 10-6. Gas extraction well design (6).
122
-------
CHAPTER 11
CASE STUDIES — RCRA/CERCLA CLOSURES
INTRODUCTION
This chapter presents five waste closure case studies.
Each study examines both design details that confirm the
suitability of the individual cap and those that may
detrimentally affect the long-term service of the facility.
The first four caps have been permitted and placed over
RCRA/CERCLA wastes. The fifth cap has been proposed
for a major municipal solid waste (MSW) landfill and
demonstrates the design problems that may be as-
sociated with such facilities.
Each design example is intended to highlight problems
that may be encountered in satisfying all aspects of
general closure criteria. The criteria that must be satisfied
are:
1. Specific minimum technology guidance (1) (MTG) ap-
plicable or appropriate and relevant to the site-
specific waste. MTG is discussed in greater detail in
Chapter 1.
2. Erosion control to limit the loss of cover soil to less
than 2 ton/acre/year, as discussed in Chapters 1 and
8.
3. Gas control systems to minimize movement of waste-
generated gases off site.
4. Ability for all systems to survive both local and global
subsidence potentials, as discussed in Chapters 1
and 2.
This chapter also raises specific concerns regarding the
use of MTG guidance blindly, without engineering confir-
mation of its suitability.
CASE 1: RCRA COMMERCIAL LANDFILL
The first closure case presents a cap over a commercial
hazardous waste disposal cell that is designed to satisfy
the basic MTG cap profile. Figure 11-1 shows details of
the general cap profile and geometry. Note that the slope
of the cap does exceed the 5 percent maximum con-
tained in the MTG criteria but is significantly flatter than
the caps on the other examples. The use of low slopes on
such facilities recognizes that the solidified waste placed
within them is very stable and will not produce significant
long-term subsidence. Such low slopes cannot be used in
applications where high long-term subsidence is a con-
cern, such as with many CERCLA and MSW closures.
This chapter examines two significant design considera-
tions for such facilities: 1) calculation of localized sub-
sidence and its impact on the cover barrier layer, and 2)
the impact of gas collection systems.
Calculation of Localized Subsidence
During the service life of this facility, it received nearly
10,000 transformers containing PCB oils (TSCA per-
mitted). The regulator expressed concern about the long-
term impact of the loss of oil and eventual collapse of the
transformer cases. Fortunately, available records
provided the location and size of the transformers. The
general subsidence model used to predict the surface
displacement of the cap due to transformer collapse was
adapted from an EPA study by Murphy and Gilbert (2)
(see Figure 11-2). The key assumption in this model is
that the volume of the surface depression is equal to the
volume of the oil leaking from the transformer. This is a
conservative assumption because it neglects the arching
that will occur within both the waste and operational soils
placed around the waste. An additional key assumption
must be made regarding the friction angle of the waste it-
self. For this case, the friction angle was assumed to be
that of the operational soils placed around the waste. For
wastes in general, such values can be measured in ac-
tual field tests (3).
The simple model for subsidence due to a single trans-
former collapse then must be applied to the actual cover
for all 10,000 transformers. The subsidences are ac-
cumulated and plotted as shown on Figure 11-3. By ex-
amining the cap's elevation contour, one can estimate
the maximum long-term relative vertical displacement of
the cap. For this case, the maximum relative displace-
ment is approximately 0.5 m in 6 m (1.8 ft in 20 ft.)
Calculation of the maximum vertical relative displacement
is important only if the designer can estimate the impact
of such displacement on the site-specific cap profile.
MTG barrier systems consist of a geomembrane and a
clay layer, both of which must be separately evaluated for
strain. The strains in the geomembrane can be estimated
using one of two models, depending on the type of an-
ticipated subsidence.
123
-------
VEGETATED TOPSOIL AND ROOT ZONE, 3 FT. MINIMUM
-SYNTHETIC DRAINAGE MEDIA,GEOTEXTILE AND GEONET
SYNTHETIC MEMBRANE (FML) BO MIL HOPE
1— COMPACTED SOIL LINER, 3 FT. MINIMUM
INTERMEDIATE COVER. G IN. MINIMUM
WASTE DETAIL OF COVER
SECTION THROUGH LANDFILL
Figure 11-1. Case 1—Cap profile and geometry.
For trench-like subsidence, the strains can be calculated
using the model shown in Figure 11-4. The maximum
strain that the geomembrane can tolerate in such a plane
strain condition is given by the uniaxial test data com-
monly reported by geomembrane manufacturers (see
Figure 11-5).
For spherical-type subsidence, the strains in the
geomembrane can be calculated using the method dis-
cussed in Chapter 3 of this manual. For such an as-
sumed failure mode, the designer must compare the
predicted strain with the ultimate strain limit of the
geomembrane, as obtained from biaxial testing (see
Figure 11-5). Chapter 3 gives additional data on typical
ultimate strains of common geomembranes in biaxial
loading. Most geomembranes can easily tolerate vertical
differential settlement of the cover in excess of 0.9 m in 3
m (3 ft in 10 ft) of run. This results in a factor of safety
based on an ultimate strain of 3.3 for the geomembrane
in Case 1.
The strain in the soil component of the barrier can be es-
timated using the chart in Figure 11 -6. The specific ul-
timate tensile strain of the onsite soil can be evaluated in
a triaxial Consolidated Isotropic Undrained (CIU) test or
can be estimated from the chart in Figure 11-7. For this
particular soil barrier, the ultimate relative strain allowable
under this criteria is 0.4 m in 3 m (1.2 ft in 10 ft) of run.
This results in a factor of safety of 1.33 for the clay com-
ponent in Case 1. If the settlements are occurring over an
extended length of time, this low factor of safety may be
acceptable due to the ability of a clay to creep. The creep
deformation of the clay allows long-term strains to
develop in the layer without a comparable increase in
stress. This is commonly referred to as "stress relaxa-
tion."
Gas Collection Systems
Commercial hazardous waste facilities generate minimal
gas due to the solidified nature of the waste. Typically,
gas collection systems for such facilities are simple linear
French drain collectors, as shown in Figure 11-8.
Extreme caution must be exercised in designing gas
removal systems for wastes that have a long anticipated
lifetime. A gas removal system is a very efficient vehicle
for surface water to gain access to the waste if the vent
pipes become damaged. Thus, if long-term maintenance
of the cap cannot be assured, a gas collection system
may eventually cause failure of the cap to perform its
primary function—preventing surface water from reaching
the waste. Provisions should be made in the permit of
124
-------
HOM2ONTAC WSPVACEMENT
STKAM
(-) I
TENSION
l\
\
- COMPRESSION -
-TENSION-
POMT OF MAXMUM
TENSU STNAM.
POMT OF MAXMUM
COMMESSVE STNAM
VERTICAL DISPLACEMENT
(SU8SOENCC CURVE)-
8s Murphy & Gilbert
J .''
MINED
Figure 11-2. Case 1—General subsidence model.
such facilrties for removing and sealing the gas vents if
postctosure monitoring indicates that no appreciable
quantities of gas are being generated.
As a final comment, the HELP analysis (see Chapters 8
and 9) for such caps must assume an effective leakage
for the geomembrane component of the barrier. This
leakage is commonly calculated by assuming from 9 to
13 penetrations (1 cm) per acre in the geomembrane.
The leakage through such penetrations can then be cal-
culated using the following equation (5):
Q = 3a°75h°-75Kd05
where Q = steady-state leakage rate
(M3/sec)
a = area of hole (M2)
h = head of leachate (M)
k = permeability of underlying soil
(MIS)
A revised version of HELP is being developed that will
accept such penetration data directly.
CASE 2: RCRA INDUSTRIAL LANDFILL
The second case study shows a cap profile that is be-
coming increasingly common in Europe and the United
States due to the high cost per acre of composite lined
landfills. As shown in Figure 11-9, the cap has two sig-
nificant profiles: a steep perimeter that provides for the
volume of the facility and a flatter top that covers the
majority of the waste. Figure 11-9 also shows a detail of
the cap profile which is a typical MTG profile. The key
design problems for this case involve the steep perimeter
of the cap, including both the sliding stability of such
covers and the erosion resistance of their protective sur-
face.
The slope stability of covers, or liner systems, containing
geosynthetic layers is typically of concern if the slope ex-
ceeds 8 degrees. The three horizontal to one vertical
(3H:1V) slopes of the perimeter are 18.4 degrees and,
therefore, of concern. The stability of cover and liner sys-
tems on such slopes is evaluated by performing
laboratory direct shear tests on each suspect interface to
determine the minimum factor of safety against sliding.
125
-------
Figure 11-3. Case 1—Cumulative subsidence.
15
J*
z
10
0£
O
i ,
z
OrcuUf
-I'
U-l
/
TrUntpilir
0 0.1 0.2 0.3
SETTLEMENT RATIO, S/2L
Figure 11-4. Case 1—Geomembrane strains in trench subsidence.
126
-------
4000
(KEORNER,RICHARDSON-UNIAXIAl)
(STEFFEN-8IAXIAU
100
200
300
400
SOO
STRAIN. %
Figure 11-5. Case 1—Uniaxlal and biaxial geomembrane response.
1 oo
0.75
0.50
0.25
100.0
Index of maximum settlement A/L vs tensile strain (after Gilbert and Murphy, 1987)
Figure 11-6. Case 1—Subsidence strain in soil barrier.
127
-------
3.5
3.0
I"
to
I ' I~T
TTT 1
LEGEND
i I I I i i
A
O
LEONARD ( BEAU FtCGJRE TESTS ) j
TSCHE30TARIOFF ( DIRECT TENSION TESTS )
WES DATA ( DIRECT TENSION TESTS )
FOR THIS STUDY (OU DIRECT TENSION TESTS)
01*
i i i i i i
10 ,00
PLASTICITY INDEX. %
strain vs plasticity mam (after G3b«ft and Mi»ony. 1987)
IOOO
Figure 11-7. Case 1—Ultimate tensile strain In clays.
This testing procedure is described in greater detail in
Chapter 3 (pp. 3-4) and Appendix A. Steep covers requir-
ing a geomembrane commonly use three liner/drainage
layer profiles to provide a stable slope:
1. A textured HOPE or VLDPE geomembrane with
either a sand drainage layer or a drainage layer
formed using a geonet with filter fabric bonded to
both faces.
2. A geomembrane having nonwoven geotextiles
bonded to both faces and a sand drainage layer.
3. A smooth geomembrane, sand, or geonet drain, with
an added geogrid reinforcement in the cover soil
layer to hold the layer on the stope.
The first two alternatives are examined for this case;
Case 3 discusses the third alternative.
Figure 11-10 shows direct shear data for the first alterna-
tive. Because the nonwoven material used in the bonded
geomembrane will develop the full shear strength of the
adjacent soil, direct shear tests are not commonly per-
formed for this material. Therefore, with both the tested
textured and bonded geomembrane, the minimal inter-
face friction angle will be significantly greater than the
18.4 degree sidestope angle. It must be noted that such
interface friction angles must be verified in laboratory
testing; not all textured geomembranes are effective.
The owner/operator must use caution in interpreting
direct shear data from evaluations of interface friction
angles. A recent full-scale field test of various cover
profiles demonstrated that the interface friction angle is
very dependent on the normaJ force acting on the layers (6).
Thus, direct shear data from tests for cap design using
tow normal toads should not be used for designing liner
interfaces where high normal toads are anticipated. This
dependency also makes the use of interface friction
angles obtained from the literature very suspect.
Laboratory direct shear tests should be performed using
the soils, geosynthetics, and normal toads associated
with the site-specific conditions.
The steep perimeter slopes must be verified as satisfying
the MTG criteria of a cover soil loss of less than 2
tons/acre/year. The rate of soil toss is verified using the
Universal Soil Loss Equation (USLE) given by:
A=RKLSCP
where K
LS
R
soil erosion factor from
Table 11-1
slope constant from
Figure 11-11
rainfall and runoff index
cover management factor
P = practice factor, 0.3 to 1.0
EPA proposed a procedure to calculate an effective LS
factor for caps having two distinct slopes, as found in this
case (7). This method, however, is not currently recom-
mended because it underestimates true soil toss. For
caps having two very distinct slopes, it is more effective
to evaluate each stope independently and to provide a
runoff collection ditch, e.g., swale, between the slopes to
hydraultoally disconnect these features in the field. Thus,
the 3H:1V perimeter slopes of Case 2 should be
evaluated using their maximum stope length and full
128
-------
Figure 11-8. Case 1—Gas collector system.
slope angle. Calculations of annual soil loss for the
3H:1V side slopes were performed using the following
values for the USLE variables:
R = 140—from local SCS
K = 0.3, sandy loam from
Table 11-1
C = .006—from local SCS
P = 1.0, maximum value
L = slope length = 15 m (50 ft)
S = slope = 3H:1V = 33 percent
These values and the LS topographic factor obtained
from Figure 11-11 yielded an annual soil loss of 1.8
tons/acre/year, which is acceptable.
Caps having two distinctive slopes may be designed
using two distinctive methods of cover protection.
Figure 11-12 shows one early scheme that incorporated
an armoring cover of coarse gravel on the steeper slope.
In this example, a swale is not provided between the
slopes due to the high transmissivity value of the gravel.
In general, however, such slopes should be separated by
a swale to make them hydraulically independent.
CASE 3: CERCLA LAGOON CLOSURE
In this case, sediments from three industrial settling
lagoons were consolidated to a single mound, as shown
in Figure 11-13. The sediments contained RCRA con-
stituents, but the age of the waste and the use of existing
lagoons for the consolidated facility made RCRA MTG
not applicable. The nature of the sediments and general
site conditions, however, made RCRA MTG appropriate
and relevant (see Chapter 1). Thus, the cap for the con-
solidated sediment mound is essentially an MTG cap
129
-------
-2'-0" MINIMUM
FINISHED GRADE OF CAP
4" MINIMUM THICK TQPSOIL LAYER
WITH VEGETATIVE COVER.
l'-8" THICK 900T ZONE EMBANKMENT
SYNTHETIC DRAINAGE MEDIA
CAP GEOMEMBRAME
._.„. 2'-0" MINIMUM THICK COMPACTED
SOIL CAP LINER
6" MINIMUM THICK FINAL INTERMEDIATE
COVER LAYER
LIMITS OF WASTE
TYPICAL CAP SECTION
SCALE: NONE
LIMITS O
TOP OF CAP
ELEV. 1220.00
SLOPE VARIES
__ (TYP.)
BOTTOM OF FINAL
INTERMEDIATE COVER
TOP OF OPERATIONAL COVER
2.78%
^— ELEV. 1185.70
SUBGRAOE Or
SOIL LINER
Figure 11-9. Case 2—Cap profile and geometry.
UJ
I
en
Normal Stress
Figure 11-10. Case 2—Direct shear data: texture HOPE.
130
-------
Table 11-1. Soil Texture Constant for Soil Loss
Evaluation
Texture class
Organic matter content
<0.5 percent
2 per cent
4 per cent
Sand
Find sand
Very fine sand
Loamy sand
Loamy fine sand
Loamy very Fine sand
Sandy loam
Fine sandy loam
Very Tine sandy loam
Loam
Silt loam
Silt
Sandy clay loam
Cay loam
Silty clay loam
Sandy clay
Silty clay
Clay
K
0.05
0.16
0.42
0.12
0.24
0.44
0.27
0.35
0.47
0.38
0.48
0.60
0.27
0.28
0.37
0.14
0.25
K
0.03
0.14
0.36
0.10
0.20
0.38
0.24
0.30
0.41
0.34
0.42
0.52
0.25
0.25
0.32
0.13
0.23
0.13-0.29
K
0.02
0.10
0.28
0.08
0.16
0.30
0.19
0.24
0.33
0.25
0.33
0.42
0.21
0.21
0.26
0.12
0.19
The values shown art estimated average! of broad ranges of specilic-soil values. When a tenure
is nearlhe borderline of two texture classes, use the average of the two K value*. For specific aoda,
use of Figure 16 or Soil Conservation Service K-value tables will provide much greater accuracy.
From ARS, 1975.
>_*
"7 ! '
• .^ -r H
1
10 20 JO 4O 60 80 IOO 300
SLOPE LENGTH , METRES
4O0600
Figure 11-11. Case 2 — Slope factors for soil loss evaluation.
~~, - j «J '.i**rr*r?rrrrrw*fr**~ "-*.".* *
J h - *'^!^* ******"'****SSSr'r'f'** r'Sr* RACVp fl_l_ "r^**
^"** ^^^f * * r r r r r f m - . ^ f* * r r r ,. O^W n r IU W* ^ ^
Figure 11-12. Case 2—Slideslope armoring scheme.
(see Figure 11-13). A key variation, however, is the use
of a commercial bentonite board in place of the com-
pacted clay component of the composite barrier system.
As in Case 2, the combination of steep slopes and
geosynthetic interfaces made slope stability a concern for
this cap. Direct shear tests of the geosynthetic interfaces
indicated that the governing interface was between the
PVC geomembrane and the geotextile forming the sur-
face water of the bentonite board. Additionally, as the
bentonite board hydrates, it loses significant shear
strength. In general, the hydrated bentonite board is not
stable on slopes steeper than 9 degrees. To ensure
stability of the drainage layer and cover soil, the design
incorporated geogrid soil reinforcement into the cap.
Chapter 3 discusses the calculations required to confirm
the ability of the geogrid. It is important that the strain as-
sumed in selecting available geogrid tensile strength be
small enough that excessive elongation of the geogrid
does not occur. The 5 percent strain assumed in this
Case 3 analysis is the maximum strain that should be
used.
131
-------
Cap Profile
DRAN&GE
LOCATION
Figure 11-13. Case 3—Cap profile and geometry.
In addition to having sufficient tensile strength, the
geogrid must be anchored sufficiently to develop this
strength. It cannot be anchored using an anchor trench
without impinging on the waste. The geogrid in Case 3,
therefore, is anchored by running the grid continuously
over the cap and counterbalancing the weights of the
cover soils on opposing faces. While this procedure is
technically simple, it restricts, construction significantly;
the cover soil and drain must be placed in a symmetrical
manner, preferably from the top down, to tension the
geogrid. Figure 11-14a shows the geogrid being placed
over the geomembrane, and Figure 11-14b shows the
drain layer being placed on top of the geogrid.
Water collected in the surface water drainage layer must
be allowed to freely leave that system to avoid building up
head on the liner, and to maintain stability. Figure 11-15a
shows the sideslope toe drainage detail used in Case 3.
From a long-term maintenance standpoint, this drainage
system is very poor. The thin layer of loam topsoil will
readily erode at the surface interface with the geotextile
and trap rock, as shown on Figure 11-15b. Surface water
drainage layers in caps having significant slopes, such as
in Case 3, should outlet using pipe laterals placed at a
minimum of one 40-cm (4-in.) drain pipe every 61 m (200
ft) around the perimeter of the surface drain.
CASE 4: CERCLA LANDFILL CLOSURE
The fourth case study is of a cover placed over an exist-
ing MSW landfill that received 20,000 yd3 (15,292 m3) of
baghouse dust containing cadmium, chromium, and lead.
The baghouse dust was placed on top of the MSW waste
and was, therefore, highly exposed. The landfill itself was
adjacent to a community park and the local youths had
established biking paths over the landfill. Both the state
and the principal responsible party (PRP) wanted to close
the landfill in a manner that prevented surface water from
reaching the dust and discouraged the recreational use
of the cap. For these reasons, they selected a unique
hardened cap profile. Such hardened caps do not
promote recreational use of the cover and, therefore, do
132
-------
Figure 11-14a. Case 3—Placement of geogrid over geomembrane.
Figure 11-14b. Case 3—Placement of drainage layer over geogrid.
133
-------
Surface Water Drainage Layer
Membrane
Zone of High Erosion
12" Trap Rock
Note: cm = in. % 2.540
Figure 11-15a. Case 3—Outlet detail for sideslope toe surface water drainage layer.
«:
Figure 11-15b. Case 3—Erosion at drainage layer outlet.
134
-------
not create an attractive nuisance in terms of maintenance
and security. Figure 11-16 shows the final cap profile and
contours.
The cap profile is significantly different than the MTG cap
in that it uses no drainage or agricultural layers. The as-
phalt and paving fabric form a unique composite barrier
with the compacted clay cap. The chip seal added to the
top of the barrier is provided to protect the asphalt and
paving fabric from ultraviolet (UV) light degradation, not
for erosion control. The "hardened" cap is advantageous
since it is not an attractive nuisance, requires very low
maintenance, and minimizes the problem of volunteer
vegetation.
The geotextile was placed over the asphalt on a surface
of asphalt emulsion (see Figure 11-17a). Rolling the
fabric over the hot emulsion fully impregnated the geotex-
tile so that it acts as a water barrier. The chip seal placed
on top of the geotextile (see Figure 11-17b) is bonded to
the geotextile by the emulsion, in a manner similar to an
industrial roofing system.
While the hardened cap is low maintenance, it does re-
quire an annual inspection and renewal of the chip seal
surface every 5 years. Additionally, the perimeter
drainage must be cleaned regularly to promote surface
water drainage. Allowable differential subsidence criteria
must be established for such caps in the same manner
as described for Case 1.
Similar hardened caps have been used on RCRA
closures in the Southeast. One particular closure at a
Department of Energy facility in Tennessee functions as
a parking lot. This particular cap replaced the agricultural
layer of the MTG profile with an asphalt and subbase
parking surface. While such caps must obviously be in-
spected on a regular basis, they can offer significant
maintenance and land use advantages.
CASE 5: MSW COMMERCIAL LANDFILL
This last example, Case 5, shows how the basic
RCRA/CERCLA closure profiles are being adapted for
the more common MSW landfills. The cap profile shown
in Figure 11-18 includes a composite barrier layer and a
protective/agricultural soil cover. It does not include a
drainage layer between the barrier layer and the cover
soil. The drainage layer is often omitted in MSW caps. In
particular, states such as New York (8) have chosen not
to require the drainage layer due to concerns regarding
Cap Profile
— 2
20' CATC
CULVERT \
Figure 11-16. Case 4—Cap profile and geometry.
135
-------
' V V ^. s- " >• x
Figure 11-17a. Case 4—Placement of geotextile on asphalt emulsion.
Figure 11-17b. Case 4—Placement of chip seal on geotextile.
136
-------
the impact of this layer on the agricultural growth placed
over the cap. It should be noted, however, that New York
requires a liner system beneath all new landfills that ac-
tually exceeds RCRA MTG criteria. Thus, the omission of
the drainage layer was not for financial reasons.
Contours for the Case 5 cap are shown in Figure 11-18
and reflect the dual slope profile developed in Case 2.
The general goals for the MSW cap are low maintenance,
minimization of infiltration, and aiding in gas collec-
tion/containment. MSW caps commonly cannot be con-
structed on new landfills in a single stage as can RCRA
and CERCLA caps. The staged construction of a MSW
cap is required because lined MSW landfills typically are
divided into adjacent cells, with each cell built to contain 4
to 6 years of waste. Figure 11-19 shows the profile of the
MSW facility with two cells having a common cover.
Facilities have been permitted with more than 10 such
cells beneath a common cover. Such facilities eliminate
the long-term exposure of the liner system that would
result if a single large cell was constructed, and do not
lose the airspace between the cells that would occur if in-
dividual covers were placed on each cell.
It is necessary to incrementally cap a facility that has
multiple adjacent cells to prevent excessive leachate
generation. Strategies for incremental cap construction
should be reviewed as part of the permitting process.
Such strategies should provide for drainage swales
spaced at intervals of no more than 6.1 m (20 ft) of verti-
cal grade change over the cap to control surface water
runoff.
MSW gas collection systems are commonly either
blanket collectors with passive vents or active systems
Cap Profile
Figure 11-18. Case 5—Cap profile and geometry.
137
-------
using discrete wells. Figure 11-20 shows the proposed
active well array for Case 5. Each well consists of a per-
forated plastic pipe within a gravel screen. The top of the
pipe will penetrate the low hydraulic conductivity barrier,
and must be sealed at the soil barrier using a bentonite
seal and at the geomembrane barrier using a boot/clamp
fixture. As the waste subsides, the gas well pipe will
move upward relative to the cap geomembrane. The
flexible boot between the pipe and the cap geomembrane
must be installed to allow such differential movements.
The boots commonly are improperly installed upside
down, e.g., they allow movement of the pipe downward
relative to the cap geomembrane. This installation,
however, must be avoided to prevent damage to the
geomembrane seal. This seal not only limits surface
water infiltration, but also aids in maintaining the low
vacuum required for active gas removal.
The use of a geomembrane in the liner and the cap will
eliminate the lateral migration of gas if the
geomembranes are intact. Perimeter gas monitoring wells
(see Figure 11-21) provide an indication of the condition
of the liner and the cap. Such wells are installed at 152-
to 305-m (500- to 1,000-ft) spacings around the perimeter
of the landfill. Most states now limit gas concentrations in
such wells to less than 25 percent of the lower explosive
limit of the methane.
CONCLUSIONS
The five case studies presented in this chapter illustrate
the need to closely evaluate the stability of closure sys-
tems related to sliding at the interfaces of the layers
making up the cap, and alternatives for controlling sur-
face erosion. Additionally, these cases highlight the fol-
lowing permit considerations:
1.
2.
The permit should contain requirements for regular
monitoring of cap subsidence, criteria for allowable
differential cap subsidence, and an agreed-upon
method for repair of excessive subsidence.
Poorly maintained gas collection systems can allow
surface water through the cap. Passive vents should
be minimized and protected from damage. Active gas
wells will move upward relative to the cap and may
damage the cap barrier. Such wells should be in-
spected regularly and removed when no longer in
production.
3.
All erosion control systems require maintenance and
regular inspection. The limits of both should be estab-
lished in the permit.
CERCLA caps in particular require careful evaluation to
determine which of the RCRA MTG cover components
are appropriate for the specific site and waste.
REFERENCES
1. U.S. EPA. 1989. Technical guidance document: final
covers on hazardous waste landfills and surface im-
poundments. EPA/530-SW-89-047. July.
2. Murphy, W.L. and P.A. Gilbert. 1987. Prediction of
landfill cover performance. Unpublished study by
COE for EPA-RREL, 1985 through 1987.
Landfill remediation. First
Municipal Solid Waste,
3. Richardson, G.N. 1990.
U.S. Conference on
Washington, DC.
4. U.S. EPA. 1988. Geosynthetic design guidance for
hazardous waste landfill cells and surface impound-
ments. EPA/600/52-87/097.
Solid Waste
, Operational Cover
/"
Geomembrane
Cell #2
Figure 11-19. Case 5—Profile showing MSW subcells.
138
-------
Well Collection System Vx%^ 11 •; ( 'SS?5**
) BLOHICK
LOCWlOH
Figure 11-20. Case 5—Gas collector well array.
8'DIA. STEEL PIPE
STEEL PIPE CAP W/
HINGE & LOCK
P.V.C.PIPE CAP
DO NOT CEMENT
CONC. BENTONITE SEAL
I1 DIA. SCH. 40 P.V.C. PIPE
W/ 3/16' DIA. (MIN.) SCREEN
HOLES
PEA GRAVEL PACK
P.V.C. END CAP
MIN. BORE DIA.
Figure 11-21. Case 5—Perimeter gas monitoring well.
139
-------
5. Bonaparte, R. et al. 1989. Rates of leakage through
landfill liners. IFAI Geosynthetics '89 Conference,
San Diego.
6. Giroud, J.P. et al. 1990. Stability of cover systems on
geomembrane covers. Proceedings, Fourth Interna-
tional Conference on Geotextiles, The Hague,
Netherlands.
7. U.S. EPA. 1980. Evaluating cover systems for solid
and hazardous waste sites. SW-867.
8. New York - 6 NYCRR Part 360.
ADDITIONAL REFERENCES
U.S. EPA. 1986. Covers for uncontrolled hazardous
waste sites. EPA/540/2-85/002.
U.S. EPA. 1985. Geotextiles for drainage, gas vent-
ing, and erosion control at hazardous waste sites.
EPA/600/2-86/085.
U.S. EPA. 1989. Requirements for hazardous waste
landfill cells and surface impoundments. Seminar
Publication. EPA/600/52-87/097.
140
-------
CHAPTER 12
POSTCLOSURE MONITORING
INTRODUCTION
The owner/operator of a facility must give significant con-
sideration during the closure permit process as to the na-
ture and extent of postclosure monitoring that will be
required. While regulatory postclosure monitoring time
frames range from 30 years for RCRA wastes to 500
years for mixed wastes (10 CFR 61), the actual monitor-
ing period will be influenced by the stability of the waste
and cover system. The permit should establish monitor-
ing procedures, acceptance criteria, and remediation
methods for the following key parameters:
1. Ground-water quality and potentiometric surface
should remain within the limits established in permit-
ting.
2. Leachate quantities and chemical makeup should
remain predictable.
3. Gas release concentrations and general air quality
must remain within guidelines. Such guidelines will
become stricter with time.
4. Differential subsidence of the cover must be limited
and repaired if allowable limits are exceeded.
5. Surface erosion must stay within the 2 ton/year allow-
able and be repaired on an annual basis.
The key elements in the monitoring program that must be
established during permitting are detection methods, al-
lowable limits, and the plan for remediation when limits
are exceeded.
GROUND-WATER MONITORING
Key monitoring variables in a comprehensive ground-
water monitoring program include both changes in the
potentiometric surface that could bring the landfill liner
system in contact with the ground water and the chemical
quality of the ground water that is an indicator of leachate
release. In RCRA facilities, both the potentiometric and
background water quality will be established during per-
mitting of the landfill prior to placement of the waste. For
CERCLA facilities, such information should be estab-
lished during the closure permit process.
A ground-water well network must be established that
both tracks changes in the ground-water potentiometric
surface and detects leakage from the facility. In both
RCRA and CERCLA facilities, the background quality of
the ground water must be documented prior to closure.
Individual monitoring wells must be designed to reflect
both the anticipated contaminant and the site-specific
stratigraphy. Figure 12-1 shows a typical well configura-
tion. The well casing will commonly be PVC for inorganic
contaminants and stainless steel for organic con-
taminants. While monitoring wells have become very
standardized, it is important to specify locking well caps
that prevent tampering of the well, well seals that restrict
surface water flow into the well, and solvent free well pipe
connections that do not contaminate the well.
In CERCLA sites, great care must be taken during the
placement of monitoring wells and during any soil borings
to avoid penetrating an aquiclude (low hydraulic conduc-
tivity soil layer) that may lie underneath contaminated
ground water. Figure 12-2 shows such a potential stratig-
raphy. When placing monitoring wells through an
aquiclude, a casing must first be installed from the
ground to the aquiclude. A grout seal is then established
to hydraulically isolate this casing from the aquiclude,
and the monitoring well is drilled to the lower aquifer
within the casing. In this manner, the contamination from
the upper aquifer will not contaminate the lower aquifer.
Ground water should be sampled at a frequency defined
by the level of anticipated contamination and the site con-
ditions. Generally, it is useful to have monthly ground-
water background data prior to permitting operation or
closure of a facility. Post-operation sampling frequency
then can be decreased to quarterly monitoring, which
should be maintained unless the consistency of measure-
ments and operation justify sampling less frequently.
Postclosure monitoring frequencies commonly range
from quarterly for lined RCRA facilities to annually for
common MSW landfills.
LEACHATE MONITORING
Both the quantity and composition of leachate generated
within a RCRA facility provide significant information on
the performance of the closure system. If the closure sys-
tem is properly designed and installed, the rate of
leachate generation in the primary collector will decrease
with time. If the closure is not complete, then the rate of
141
-------
1 2 3
1 2
1 2
Sand (K = 1 x 10~3 cm/sec)
Layer3
Figure 12-1. Monitoring well configuration.
8'00 LOCKING PROTECTIVE
STEEL CAP
SIDE VENTED PVC WELL CASING CAP
LOCK
7'to STEEL PROTECTIVE CASING
5VY LONG
V SCM. <0 PVC RISER PIPE
ISTICK-UPI
ORAINHOLE
ORIGINAL CSOUNO SURFACE
REINFORCED CONCRETE CAP —
UIN. 21 RADIUS '»/ •« REBAR
ON 61CENTERS
CONCRETE PLUG EXTENDING 36'
DOWN BOREHOLE BELOW CAP
4- SCH. to PVC FLUSH JOINT CASINO
SCREENED INTERVAL 2'-20
NORMALLY
VARIABLE CUP LENGTH.
NORMALLY 6'-r
CEMENT/BENTONITE GROUT fLACEO
BY SIDE DISCHARGE TREMIE PIPEi
« LBS.PORTLAND CEMENT
5 LBS. POWDERED 3ENTOMTE
6 GALS. WAFER
ILB. CALCIUM CHLORIDE
BENTOMIE ftLLET SEAL-
TAUPED AND HYDRA TED
FINE SANO FILTER
SAND PACK. CONSISTING OF
WASHED I GRADED SILICA SANO.
SIZED FOR THE AQUIFER AND
PLACED BY TREMIE PIPE
FACTORY SLOTTED OR CONTNUOUS WIRE
SCREEN SIZED FOR AQUIFER GRAIN
SIZE DISTRIBUTION
TAILPIECE OR SEDIMENT CUP
CENTRALIZEP. (SPACED AS REO'O.)
PVC END CAP (THREADED)
BOTTOM OF BOREHOLE
Figure 12-2. Monitoring interbedded aquifer.
142
-------
Impact of Biological Growth
CD
<
LLJ
2
cc
LLJ
Q.
BIOCIOC BXCXfLOSS
II
BIOCICE BACKrLOSH
12
1 5 T
3 0 T
TIME
Figure 12-3. Impact of biological growth on filters.
loo-i
20 40 60 80 100 120
5L Rate
5L Acids
Figure 12-4. Gas generation versus time.
leachate generation in the primary collector may reflect
precipitation trends. Therefore, the integrity of the closure
can be verified by evaluating leachate quantity records. A
sudden increase in the quantity of leachate generated will
clearly indicate failure of the closure. Unfortunately, many
CERCLA facilities will lack a liner and primary collector
system.
The concentration of contaminants in a facility's leachate
will increase with time until an equilibrium condition is es-
tablished. A sudden reduction in this level of con-
taminants is a good indication that the cover has been
breached, allowing a slug of surface water to enter the
waste and dilute the leachate. Biological growth,
however, can also have a significant impact on the
monitoring system over the long term. Figure 12-3 shows
the impact of biological growth in municipal solid waste
(MSW) leachate on the permeability of a geotextiie filter
commonly used in collector systems (1). The time, T, for
significant reduction in permeability may be as short as 6
weeks. Thus, a long-term decrease in the amount of
leachate generated may indicate biological clogging of
the collector, which may prevent detecting failure of the
closure. Such biological clogging occurred recently in a
MSW landfill in Delaware. A significant head of perched
leachate was discovered within the waste while the quan-
tity of leachate generated was actually decreasing. This
clogging required excavation of the waste and replace-
ment of the primary collector system.
GAS GENERATION
Gas generation within a waste containment system must
be monitored both to ensure that such gas does not
143
-------
Table 12-1. Threshold Limits of Air Contamination
Contaminant
Threshold Limit Values of Selected Air Contaminants"
TLV
Dust
Carbon monoxide
Asbestos
Benzene
Coal dust
Cotton dust
Grain dust
Hydrogen sulfide
Nuisance particulates
Phenol
Vinyl chloride
Wood dust
Hard wood
Soft wood
I mg/m3
50 ppm
0.2 to 2 fibers/cm3 (depending on asbestos type)
10 ppm
2 mg/mj
0.2 mg/m3
4 mg/m3
10 ppm
10 mg/m3
5 ppm
5 ppm
I mg/m3
5 mg/m3
"Values of TLV obtained from the American Conference of Governmental Industrial
Hygienists (I987).
migrate off site and to indicate closure performance. The
rates of gas generation vary from more than 900 liters/kg
waste/year in MSW wastes (2) to insignificant rates in
RCRA commercial landfills. The rate of gas generation in
future MSW landfills is anticipated to decrease as these
landfills are constructed with liners and leachate collec-
tion systems. The addition of a geomembrane in the
cover will significantly decrease the amount of surface
water infiltration and also lead to lower gas generation
rates.
When geomembranes are used in a cover, very little gas
can escape vertically. Therefore, in an unlined facility,
such as a typical CERCLA closure, escaping gas will
move to the perimeter of the cover. Simple gas monitor-
ing wells (described in Chapter 11, Case 5) must be in-
stalled around the perimeter of the cover to detect
laterally moving gas. The level of gas at such wells must
remain below 25 percent of the lower explosive limit
(LEL). The level of gas production can vary significantly
with the weather; therefore, the monitoring frequency
should be increased when the surrounding ground is
saturated or frozen.
Gas odors detected above a closure system that includes
a geomembrane indicate that the geomembrane has a
significant penetration. A regular survey of gas levels on
the surface of the closure is a good method of verifying
the integrity of the cap barrier.
As detailed in Chapter 11, gas removal systems must be
designed with a minimal number of penetrations through
the cover system. Each vent is a potential major leak. For
passive systems, a maximum of one vent per acre should
be included initially. If monitoring of these vents reveals
excessively high gas concentrations, then additional
wells can be installed. In active systems, gas wells must
be removed when they are no longer productive to
prevent damage to the cover.
As with leachate quantities, the rate of gas generation
should also decrease with time if the cover system is
functioning properly such that moisture does not reach
the waste. Figure 12-4 shows the result of laboratory
column gas generation tests (3). In the figure, methane
production rate and the level of carboxylic acids in the
leachate decrease with time. A properly functioning cover
will ensure that the leachate will remain acidic and that
gas production will be low.
SUBSIDENCE MONITORING
Chapter 11 discusses the ability of the cap barrier com-
ponents to tolerate differential settlements due to waste
subsidence. In Case 1, differential settlements as large
as 0.5 m in 6 m (1.8 ft in 20 ft) were tolerated by com-
posite barriers. Thus, the level of differential settlements
of interest during postclosure monitoring can be quite
large. Such levels can commonly be found by walking the
cover after a rain storm and looking for major puddles or
ponding. Subsidence depressions also can be found
through an annual survey of the cover using either con-
ventional or aerial survey methods.
Subsidence depressions must be remediated below the
level of the barrier system to avoid long-term acceleration
of the subsidence due to a "roof ponding" mechanism.
Roof ponding refers to the common failure in flat roof sys-
tems 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. Remediation requires removing the cover
system in the region of subsidence and backfilling the
depression with lightweight fills. This fill may either be
144
-------
more waste or commercial lightweight aggregates. The
full cover profile must then be rebuilt over the new fill.
SURFACE EROSION
All cover systems will erode and require long-term main-
tenance. Cover systems with moderate slopes and an
agricultural cover will typically require annual main-
tenance of 0.5 percent of their surface area; this percent-
age increases with slope. Thus, all covers that use
agricultural covers require an annual inspection and
repair program. Such repair may include cleaning out sur-
face water swales, replacing cover soil, and reestablish-
ing vegetation. Areas of the cover requiring repeated
repair may benefit from hardening or the use of geosyn-
thetic erosion control blankets. Covers that use hardened
erosion control systems should also be inspected annual-
ly, though annual maintenance should not be required.
The annual inspection should verify that the agricultural
cover is being mowed at least annually to prevent the
growth of deep-rooted volunteer vegetation. In arid
regions of the country or during droughts, full RCRA
covers may not be able to maintain vegetation unless the
plants are very drought resistant. This loss of vegetation
is due to moisture loss in the root zone of the cover soil,
resulting from the underlying drainage system.
AIR QUALITY MONITORING
Air emissions from waste storage facilities will come
under increasing scrutiny in the next decade. Monitoring
techniques will be similar to those used at industrial
facilities and include passive samples obtained using col-
lection media, grab samples obtained in evacuated
sample vessels, and active pump and filter samples. The
most common air contaminants coming from the waste
disposal cell obviously are waste dependent; for MSW
wastes, these are methane, vinyl chloride, and benzene.
Table 12-1 presents typical allowable limits of selected air
contaminants. Such limits are currently undergoing exten-
sive review; significantly lower allowable levels are an-
ticipated for future operations.
The geomembrane component of the MTG cover com-
posite barrier system controls air emissions significantly.
In fact, the presence of emissions indicates that the
geomembrane cover has failed and needs to be repaired
immediately.
REFERENCES
1. Unpublished research at Geosynthetic Research In-
stitute, personal communication with R.M. Koerner.
2. Walsh, J.L. 1988. Handbook on biogas utilization.
Georgia Tech Research. February.
3. Barlaz, M.A. et al. 1989. Bacterial population
development and chemical characteristics of refuse
decomposition in a simulated sanitary landfill. Applied
and Environmental Microbiology. January.
145
-------
APPENDIX A
STABILITY AND TENSION CONSIDERATIONS REGARDING COVER SOILS ON
GEOMEMBRANE-LINED SLOPES
-------
Stability and Tension Considerations Regarding Cover Soils on
Geomembrane Lined Slopes
by
Robert M. Koerner and Bao-Lln Hwu
Geosynthetlc Research Institute
Drexel University
Philadelphia, Pennsylvania 19104
Abstract
The occurrence of cover soil instability in the form of sliding on geomembranes is far too
frequent. Additionally, there have been cases of wide width tension failures of the underlying
geomembranes when the friction created by the cover soil becomes excessive. While there are
procedures available in the literature regarding rational design of those topics, it is felt that a
unified step-by-step perspective might be worthwhile. It is in this light that this paper is
written. Included are four separate, but closely interrelated, design models. They are the
following;
• cover soil stability on side slopes when placed above a geomembrane,
• cover soil reinforcement provided by either geogrids or geotextiles.
• wide width tension mobilized in the geomembrane caused by the interface friction of
the soils placed above and below the geomembrane, and
• circumferential tension mobilized in the geomembrane by subsidence of the subgrade
material beneath the geomembrane.
Each of these designs are developed In detail and a numeric problem is framed to illustrate the
design procedure. Emphasized throughout the paper is the need for realistic laboratory test
values of interface friction, in-plane tension and out-of-plane tension of the geomembranes.
By having realistic experimental values of allowable strength they can be compared to the
required, or design, strength for calculation of the resulting factor-of-safety against
instability or failure.
A-l
-------
Stability and Tension Considerations Regarding Cover Soils on
Geomembrane Lined Slopes
Introduction
Geomembrane lined soil slopes are common in many areas of civil engineering
construction but nowhere are they more prevalent than in the environmental related field of
the containment of solid waste. Cover soils on geomembranes placed above the waste as in
landfill caps and closures as well as lined side slopes beneath the waste are commonplace as
the sketches of Figure 1 indicate. The variations of soil types beneath the geomembrane as well
as above the geomembrane are enormous. They range from moist clays in the form of
composite liners to drainage sands and gravels of very high permeability. The likelihood of
having other geosynthetlc materials adjacent to the geomembranes (like geotextlles, geonets
and drainage geocomposltes) presents another set of variables to be considered. Lastly, the
existence of many different geomembrane types, having different thicknesses, strengths.
elongations and surface characteristics leads to the necessity of performing a rational design
on such systems. Clearly, the development of design models to evaluate the stability of the
overlying materials as well as the tensile stresses that may be Induced In the underlying
geomembranes should always be performed. Fortunately, both the stability of the overlying
soil materials and the reduction of tensile stress in the geomembrane can be accommodated by
reinforcing the cover soil with either geogrlds or geotextiles. This is becoming known as
veneer stability reinforcement and is necessitated due to a number of cover soil stability, or
sloughing, failures, some of which are shown in the photographs of Figure 2.
This paper presents several design models and their development Into design equations
for cover soil stability (both without and then with reinforcement) and for the induced tensile
stresses that are mobilized In the underlying geomembrane. The approach taken In this paper
will utilize a single geomembrane, but It should be recognized that double liners are frequently
used beneath solid and liquid waste. Design considerations into the secondary liner, however.
can be handled by reasonable extensions of the material to be presented.
A-2
-------
Geomembrane
Waste
SOIL
SUB-SOIL
COVERS
SUB-SOIL
"Geotextile
Jeonet or Geocomposite
'Geomembrane
Waste'
(a) Landfill Cover with Soil
Above Geomembrane
(b) Landfill Cover with Drainage
Geosynthetic Above Geomembrane
Proposed,
" Waste'
Geomembrane
SUB-SOIL
(c) Landfill Liner with Soil
Above Geomembrane
Geotextile
Geonet or'
Geocomposite
Geomembrane-
SUB-SOU
(d) Landfill Liner with Drainage
Geosynthetic Above Geomembrane
Figure 1 - Various Solid Waste Geomembrane Covers and Liners Involving
Natural Soils and/or Drainage Geosynthetics
A-3
-------
Figure 2 - Cases of Cover Soil Instability for Case l(a) (upper photo) and for Case l(c) (lower
photo) as shown In Figure 1
A-4
-------
Interface Friction Considerations
It will be seen that at the heart of the design equations to be developed in this paper are
interface friction values between the geomembrane and the overlying soils or drainage
geosynthetics and also against the underlying soils or drainage geosynthetics. These values
are obtained by direct shear evaluation in simulated laboratory tests. Unfortunately, many
aspects of the direct shear test have not yet been standardized (although ASTM has a Task
Group working on a draft Standard), and many important details must be left to the design
engineer and testing organization. For example, the following items need to be carefully
considered.
minimum or maximum size of shear box
aspect dimensions of the test specimen
type of fixity of the geomembrane to the shear box and to a substrate
moisture conditions during normal stress application
type of liquid to use during sample preparation and testing
method and duration of normal stress application
strain controlled or stress controlled shear application
rate of shear application
moisture conditions and drainage during shear application
duration of test
number of replicate tests at different normal stresses
linearity of resulting failure envelope
Thus the use of reported values in the published literature can only be used with considerable
caution and, at best, for preliminary design.^1"3) For final design and/or permitting, the site
specific conditions and the proposed materials must be used in the tests so as to obtain realistic
values of the shear strength parameters adhesion (ca) and Interface friction (5), Additionally,
tests should also be performed on the soil by Itself so as to obtain a reference value for
comparison to the inclusion of the geomembrane. Calculation of the adhesion efficiency on
soil cohesion and a frictional efficiency to that of the soil by itself are meaningful In assessing
the numeric results of the designs to follow. I.e.
Ec = ca/c(100) (1)
E = tan5/tan(100) (2)
0
where
Ec = efficiency on cohesion
ca = adhesion of soil-to-geomembrane
c = cohesion of soil-to-soil
E. = efficiency on friction
8 = friction angle of soil-to-geomembrane
<)> = friction angle of soil-to-soil
A-5
-------
Past Investigations and Analyses
The Isolation of free body diagrams depicting the site specific situation to be analyzed is
certainly not new. It is a direct extension of geotechnical engineering of soil stability and is
reasonably straightforward since the failure plane against the geomembrane is clearly
defined. Thus a computer search is generally not necessary to locate the minimum factor-of-
safety stability value. Also it should be recognized that the failure surface is usually linear,
rather than circular, log spiral, or other complicated geometric shape in that it follows the
surface of the geomembrane itself.
A procedure which nicely accommodates a clearly defined straight line slip surface has
been developed by the U.S. Army Corps of Engineers.M Their wedge analysis procedures form
the essence of the developments to follow. The graphic procedures are outlined in Reference #4
but are developed in this paper into design equations in a more rigorous manner. Also to be
mentioned is the work of Giroud and Beech^5) and Giroud, et al.^ in providing excellent
insight into several aspects of the design.
In the first referenced paper, by Giroud and Beech^, a two-part wedge method is utilized
to arrive at a similar equation as ours except without an adhesion term. Also, the treatment at
the top of the slope is slightly different. Their work will be referenced in the second problem of
this set of four examples and a comparison of results will be made. In the second referenced
paper, by Giroud, et al.^, a large overburden stress necessitated the use of arching theory to
recognize that a limiting value will occur when the geomembrane is located beneath deep fills.
This is not the case with the shallow overburden stresses Imposed by cover soils placed on
geomembranes that are the focus of this paper. In the fourth example to be presented we will
use the full thickness of the overburden times Its unit weight. Additionally, we will not use a
deformation/strain reduction value in the interest of being conservative. The reference cited
by Giroud, et aU9) should be used In this regard.
A-6
-------
Model #1: Stability of Cover Soil Above a Geomembrane
Consider a cover soil (usually a permeable soil like gravel, sand or silt) placed directly
on a geomembrane at a slope angle of "CD". Two discrete zones can be visualized as seen in
Figure 3. Here one sees a small passive wedge resisting a long, thin active wedge extending the
length of the slope. It is assumed that the cover soil is a uniform thickness and constant unit
weight. At the top of the slope, or at an Intermediate berm, a tension crack in the cover soil is
considered to occur thereby breaking communication with additional cover soil at higher
elevations.
Resisting the tendency for the cover soil to slide is the adhesion and/or interface
friction of the cover soil to the specific type of underlying geomembrane. The values of "ca"
and "6" must be obtained from a simulated laboratory direct shear test as described earlier.
Note that the passive wedge Is assumed to move on the underlying cover soil so that the shear
parameters "c" and "<)>". which come from soil-to-soil friction tests, will also be required.
PASSIVE
WEDGED
WP
A ACTIVE
WEDGED r- Cover Soi,
Geomembrane
Figure 3 - Cross Section of Cover Soil on a Geomembrane Illustrating the Various Forces
Involved on the Active and Passive Wedges
A-7
-------
By taking free bodies of the passive and active wedges with the appropriate forces being
applied, the following formulation for the stability factor-of-safety results, see Equation 3.
Note that the equation is not an explicit solution for the factor-of-safety (FS), and must be
solved indirectly. The complete development of the equation is given In Appendix "A".
(FS)2 [0.5 y LH sin2 (2 co)] - (FS) |y LH cos2 co tan 8 sin (2 co) + c a L cos co sin (2 co)
+Y LH sin2 co tan <)> sin (2 co) + 2 c H cos co +yH 2tan ]
+ [(yLH cos co tan 8 + caL) (tan Q> sin co sin (2 co)] = 0 (3)
Using ax2 + bx + c = 0, where
a = 0.5 Y LH sin22co
b = -[ Y LH cos2co tan 8 sin (2co) + caL cos co sin (2 co)
+ Y LH sin2co tan <(> sin (2co) + 2cH cos co +Y H2 tan $]
c = (Y LH cos co tan 8 + caL) (tan sin co sin (2co))
the resulting factor-of-safety is as follows:
When the calculated factor-of-safety value falls below 1.0. a stability failure of the cover soil
sliding on the geomembrane is to be anticipated. However, it should be recognized that seepage
forces, seismic forces and construction placement forces have not been considered in this
analysis and all of these phenomena tend to lower the factor-of-safety. Thus a value of greater
than 1.0 should be targeted as being the minimum acceptable factor-of-safety. An example
problem illustrating the use of the above equations follows:
Example Problem: Given a soil cover soil slope of co = 18.4° (I.e.. 3 to 1),
L = 300 ft.. H = 3.0 ft. Y = 120 lb/ft3. c = 300 lb/ft2. ca = 0.4 = 32°. 8= 14°,
determine the resulting factor-of-safety
Solution:
a =.0.5 (120) (300) (3) sin2 (36.8°)
= 19.400 lb/ft
b =- [(120) (300) (3) cos2 (18.4°) tan (14°) sin (36.8°)
+ 0 + (120) (300) (3) sin2 (18.4°) tan (32°) sin (36.8°)
+ 2 (300) (3) cos (18.4°) + 120 (9) tan (32°)1
A-8
-------
= - [14523 + 0 + 4028 + 1708 + 675]
= - 20.934 Ib/ft
c = [(120) (300) (3) cos (18.4°) tan (14°) + 0]
[tan (32°) sin (18.4°) sin (36.8°)]
= [25500] [0.118]
= 3019 Ib/ft
20.934 + V (-20934) - 4(19400) (3019)
= 38,800
FS = 0.91, which signifies that a
stability failure will occur
A-9
-------
Model #2: Reinforcement of Cover Soil on a Geomembrane
Once the cover soil factor-of-safety becomes unacceptably low for the site specific
conditions (as illustrated In the previous problem), a possible solution to the situation is to add
a layer of geogrid or geotextile reinforcement as shown in Figure 4. In the case of landfill
covers, the tensile stresses that are mobilized In the reinforcement are carried over the crown
to (generally) an equal and opposite reaction on the opposing slope. Alternatively, these
stresses can be carried in friction via an anchorage mode of resistance as would occur in an
intermediate berm situation. For a landfill liner, the stresses in the reinforcement are
generally carried to an individual anchor trench extending behind the geomembrane anchor
trench. If the reinforcement is a geogrid it Is placed within the cover soil so that soil can
strike-through the apertures and the maximum amount of anchorage against the transverse
ribs can be mobilized. When using geotextiles, they can be placed directly on the geomembrane.
or embedded within the cover soil so as to mobilize friction in both surfaces.
The tensile stress of the reinforcement layer per unit width is calculated by setting "E^"
equal to "Ep" in Figure 3 and solving for the unbalanced force T" in Figure 4 which Is required
for a factor-of-safety equal to one. This value of T becomes Trec.^ which is given in Equation 5.
The complete development is available in Appendix "B".
yLH sin (o>-8)
cos <{>
cH
. .
+— - - tan <(>
sin co sin 2co
rn __ I ^ \*"* __;__ — T _
re + co)
(5)
This value is now compared to the allowable wide width tensile strength of the particular
geogrid or geotextile under consideration, i.e.,
FS=Tallow/Treqd (6)
Note that the value of "Tallow" must Include such considerations as Installation damage, creep
and long-term degradation from chemical or biological interactions. If the value is obtained
from a test such as ASTM D-4595. the wide width strip tensile test, the use of partial factors-of-
safety is recommended to accommodate the above items. W An example problem using
Equations 5 and 6 follows:
Example Problem: Continue the previous problem of cover soil
instability where a geogrid with allowable wide width tensile strength of
4000 Ib/ft is being considered (I.e., the value Includes the above
mentioned partial factors-of-safety). What Is the resulting overall factor-
of-safety? The parameters are co = 18.4°, L = 300 ft., H = 3.0 ft..
A-10
-------
Reinforcement
•T (Geogrid or
Geotextile)
T (Geogrid or
iGeotextile)
Proposed
Waste
Geomembrane
COVER SOIL
Reinforcement
SUB-SOIL
Figure ,4 - Geogrid or Geotextile Reinforcement of a Cover Soil Above
Waste and of a Cover Soil on'a Geomembrane Beneath Waste
A-ll
-------
Y =120 Ib/ft3, c = 300 Ib/ft2, ca = 0.(|>= 32°. 8= 14'
Solution:
(120) (300) (3) sin (4.4°)
COS (14°
f (300) (3) (120) (9)
cos <32°> [sin (18.4°) + sin (36.8°) ta"
cos (50.4°)
= 8539-0-5292
Treqd = 3247 Ib/ft
FS = Tallow /Treqd
_ 4000
= 3247
FS = 1.23, which is marginally acceptable and a
stronger reinforcement or a double layer should be
considered.
Note: Using the formulation developed by Giroud and Beech^5) with
the soil cohesion equal to zero results in a Treqcj = 6890 Ib/ft. while
the above formulation adjusted for a zero cohesion results in Treqcj =
7040 Ib/ft. Thus the methods appear to be comparable to one
another.
A-12
-------
Model #3: Geomembrane Tension Stresv.:., OHO i;. Unbalanced Friction Forces
The shear stresses from the cover soil above the liner act downward on the underlying
geomembrane and in so doing mobilize upward shear stresses beneath the geomembrane from
the underlying soil. The situation is shown in the sketch of Figure 5,
T (in Geomembrane)
where
u = cau+(Wcosco) tan8u
= caL +(W cosco) tan §L
Figure 5 - Shear and Tensile Stresses Acting on a Covered Geomembrane
Here three different scenarios can be envisioned:
=TL. t*16 geomembrane goes into a state of pure shear which should not be of great
concern for most types of geomembranes
• If TTJ t L. the geomembrane goes into a state of pure shear equal to TL and the balance
of Tu-TLmust be carried by the geomembrane in tension.
This latter case Is the focus of this part of the design process. The situation generally occurs
when a material with high interface friction (like sand or gravel) Is placed above the
geomembrane and a material with low interface friction (like high moisture content clay) Is
placed beneath the geomembrane. The essential equation for the design is as follows where "T"
is In units offeree per unit width, I.e., T/W. The complete derivation follows in Appendix "C".
T/W = [(caU - CjjJ + yH cos co (tan fy - tan 5j] L
(7)
A-13
-------
The resulting value of force per unit width "T/W is then compared to the allowable strength of
the geomembrane which is shown schematically for different geomembranes in Figure 6. The
target values are Tbrcakfor scrim reinforced geomembranes, Ty^y for semi-crystalline
geomembranes and Tauow (at a certain value of strain) for nonreinforced flexible
geomembranes. Note that these curves should be obtained from a wide width tensile test which
is currently under development in Committee D-35 on Geosynthetics
break
FORCE
WIDTH
yield
allo
CSPE-R
CPE-R
,EIA-R
e - 20 to 50%
STRAIN
Figure 6 - Tensile Behavior of Various Geomembrane Types
Since there is generally no reduction for partial factors-of-safety In these values of laboratory
obtained strength, the final factor-of-safety In the design should be quite conservative. An
example problem follows:
Example problem: Given the same landfill cover as described In the
previous problems with a geomembrane having an allowable strength of
2000 Ib/ft. The shear strength parameters of the geomembrane to the
upper soil are cay = 0 Ib/ft2 and 8jj = 14° and to the lower soil are 0^= 50
Ib/ft2 and 6L= 5°. Calculate the tension in the geomembrane and the
A-14
-------
resulting factor-of-safety against geomembrane failure.
Solution:
T/W = [(cau - caj + 7H cos co (tan % - tan 5j] L
= [(0-50) + (120) (3) cos (18.4°) (tan (14°) - tan (5°) 1J300
= [-50 + 55.3] 300
= 15901b/ft
FS = Tallow/Treqd
2000
~ 1590
FS = 1.25, which is barely acceptable.
Note: An alternative design to the above is to bench the cover soil
(thereby decreasing the slope length) or use a liner whose lower surface
has a higher adhesion or a higher friction surface, than the one used in
the example thereby increasing "caL" and/or "5^".
A-15
-------
Model #4: Geomembrane Tension Stresses Due to Subsidence
Whenever subsidence occurs beneath a geomembrane and It Is supporting a cover soil
some induced tensile stresses will occur due to out-of-plane forces from the overburden. Such
subsidence is actually to be expected in closure situations above completed or abandoned
landfills where the underlying waste is generally poorly compacted. The magnitude of the
induced tensile stresses in the geomembrane depends upon the dimensions of the subsidence
zone and on the cover soil properties.
The general scheme is shown In Figure 7 where the critical assumption is the shape of
the deformed geomembrane. In the analysis which is provided In Appendix "D", the deformed
shape is that of a spheroid of gradually decreasing center point along the symmetric axis of the
deformed geomembrane.^ As a worst case assumption, the geomembrane is assumed to be
fixed at the circumference of the subsidence zone. The required tensile force In the
geomembrane can be solved In terms of a force per unit width "Treqcj", or as a stress. l.e. "areqd"-
The latter will be used In this analysis since it will be compared to a laboratory test method
resulting in the compatible term. The necessary design equation Is as follows where the
specific terms are given In Figure 7.
Jreqd
cs
(8)
Cover Soil
Unit Weight
D
Geomembrane
Figure 7 - Tensile Stresses in a Geomembrane Mobilized by Cover Soil and Caused by
Subsidence
A-16
-------
Upon calculating the value of areqci for the site specific situation under consideration, it is
compared to an appropriate laboratory simulation test. Recommended at this time is a three-
dimensional, out-of-plane, tension test of the same configuration as Figure 6. It is available as
GRI Test Method GM-4J10> Thus the formulation for the final factor-of-safety becomes the
following:
Since the value of
-------
Summary and Conclusions
The occurrence of cover soils sliding off geomembrane lined slopes Is not an Infrequent
incident. While less obvious, but of even greater concern, there are often tensile stresses
imposed on the underlying geomembrane. The occurrence of extensive tensile failures of
geomembranes on side slopes is also known to have occurred.. This paper is focused toward a
series of four design models to be used to analyze various aspects of the situation.
The first model considered the cover soil's stability by itself. The design procedure is
straightforward but it does require a set of carefully generated direct shear tests to realistically
obtain the interface friction parameters.
The growing tendency toward steeper and longer slope angles gives rise to the second
design model which Is veneer reinforcement of the cover soil. Geogrids and geotextlles have
shown that they can nicely reinforce the cover soil and the first design example was modified
accordingly. The design leads to the calculation of the required tensile strength of the
reinforcement. This value must then be compared to a laboratory generated wide width tensile
strength of the candidate reinforcement material. It is important In this regard to consider
long term implications which can be addressed by partial factors-of-safety.
Both of the above analyses serve to set up the third design scenario, that being a
calculation procedure for determination of the induced tensile stresses in the underlying
geomembrane brought on by unbalanced friction values. Whenever the frictional
characteristics beneath the liner are low (e.g., when the liner is placed on a high moisture clay
soil as It Is In a composite liner), this type of analysis should be performed. The tensile stress
in the geomembrane is then compared to the wide width tensile strength of the geomembrane
for Its resulting factor-of-safety.
Lastly, a design procedure for calculation of out-of-plane generated tensile stresses in
the geomembrane was developed. This situation could readily arise by subsidence of solid
waste beneath the geomembrane. The resulting tensile stresses In the geomembrane must then
be compared to a properly simulated laboratory test for the factor-of-safety. Such three
dimensional axi-symmetric test procedures are currently available.
Each of the four above described models along with their design/analysis procedures
were illustrated by means of an example problem dealing with a cover soil in a solid waste
closure situation. This type of application is the primary focus of the paper. However, similar
situations can arise elsewhere. For example, the same situation occurs in the case of gravel
covered primary geomembrane liners on the side slopes of unfilled, or partially filled,
landfills. These slopes may have to be exposed to the elements for many years until the waste
A-18
-------
provides sufficient passive resistance and final stability. In the meantime, cover soil
instability will cause sloughing and can expose the geomembrane to ultraviolet light, high
temperatures via direct exposure, and a significantly shortened lifetime.
Hopefully, the use of design models such as presently here (and elsewhere), coupled with
the appropriate test method simulating actual field behavior, will lead to recognition of the
problems encountered and to a widespread rational design of cover soils on geomembrane
lined side slopes.
A-19
-------
References
1. Martin. J. P.. Koerner, R. M. and Whitty, J. E., "Experimental Friction Evaluation of
Slippage Between Geomembranes, Geotextiles and Soils," Proc. Intl. Conf. on
Geomembranes, IFAI, Denver, CO, 1984, pp. 191-196.
2. Koerner, R. M., Martin, J. P. and Koerner, G. R., "Shear Strength Parameters Between
Geomembranes and Cohesive Soils," Jour. Geotex. and Geomem., Vol. 4, 1986. pp. 21-30.
3. Mitchell. J. K.. Seed, R. B. and Seed. H. B.. "Kettleman Hills Waste Landfill Slope Failure. I
- Liner-System Properties," Jour. Geotechnical Engineering. Vol. 116, No. 4. April 1990.
pp. 647-660.
4. .Manual EM 1110-1902, U. S. Army Corps of Engineers. Washington. DC. 1960.
5. Giroud. J. P. and Beech, J. F., "Stability of Soil Layers on Geosynthetlc Lining Systems."
Geosynthetics '89 Conference, San Diego, CA. IFAI. 1989. pp. 35-46.
6. Giroud, J. P.. Bonaparte, R.. Beech. J. F. and Gross, B. A., "Design of Soil Layer -
Geosynthetic Systems Overlying Voids," Journal of Geotextiles and Geomembranes, Vol. 9,
No. 1. 1990, Elsevier.pp. 11-50.
7. Koerner. R. M.. Designing with Geosvnthetlcs. 2nd Edition. Prentice Hall Publ. Co..
Englewood Cliffs. NJ, 1990. 652 pgs.
8. , Standard Test Method for Determining Performance Strength of Geomembranes by
the Wide Strip Tensile Method." ASTM Draft Designation D35.10.86.02 (in task group
status).
9. Koerner, R M.. Koerner. G. R. and Hwu. B-L, "Three Dimensional, Axi-Symmetric
Geomembrane Tension Test." Geosynthetic Testing for Waste Containment Applications,
ASTM STP 1081. Robert M. Koerner, Editor ASTM. Philadelphia. PA, 1990.
10. , GRI Test Method GM-4. Three Dimensional Geomembrane Tension Test,"
Geosynthetic Research Institute. Philadelphia PA. 1989.
A-20
-------
Appendix "A"
Derivation of FS for Cover Soil Stability on a Geomembrane
Active
Wedge
W,
Passive
Wedge
Passive Wedge
H:
Y H
sin co cos co
'P ^
H
COS CO
sm 2 co
C H
FS sin co
A-21
-------
Passive Wedge
EF = WP
DE EF
WP
sin(90°-<))D)
DE = WP • tan
AD
sin (90° + <|>D) sin(90°-<|)D-i
Ep Cp + Wp • tan D
COS (}>£) ~ COS (<(>D + 05)
EP =
cos <)>£)• [Cp 4- Wp • tan <)>D]
COS<|>D
cos (D +1
[ C H
' L FS sin i
_ .
05 sin
cos ((j +
• tan
(cos <|>D cos 05 - sin <{>D sin
1
(cos 05 - tan <)>D sin
FS
[ C-H
' L FS • sin
y-H2 tan$
05 2 sin 05 cos 05 FS
r2-C-H-cosQ5 + Y'H2-
|_ 2 • sin 05 • cos 05 • FS
(FS • cos 05 - tan (|) • sin
2 • sin 05 • cos 05 • FS
A-22
-------
Active Wedge
WA = Y L H
90+®
co-8D
sin (co - 8j-,) sin (90° + SQ)
WA sin (co - 8D)
FJ « — ^^^^^^^^~^^^^^^^^^^«^^« — *>A
*\ f>f^ft Si **
WA
cos 8D
cos 8D
w fsin co cos SD - cos co sin 8D 1
cos8D
90+w
H (sin co- cos co tan SD) —
FS
A-23
-------
EA-EP
T TT , . - x ca' L (2-C-H-COSG5 + Y- H2- tan)
Y • L • H • (sin 05 - cos 03 • tan 5n) -- ^r~ = TF? - — - r — r— -r — , . ,T
' v u FS (FS • cos 05 - tan • sin 03) • (sin 2 0
H
Y'L-H-
H f ' -
H- ^sm03-
cos g ' tan S ^ - C* L _ (2 ' C • H • cos 03 + y • H2 • tan <}>)
- -
ps
ps
cos 03 -tan <(>• sin 03)- (sin 2
Y-L-H- (sin 03 • FS - cos 03 • tan6)-Ca-L (2 • C • H • cos 03 + y • H2 • tanfr)
FS
(FS • cos 03 • sin 203 - tan <(> • sin 03 • sin 203)
Y• L • H • sin 03 • FS -y • L • H • cos 03 • tan 8 - Ca • L _ (2 • C • H • cos 03 + y -H2 • tan <]>)
FS
(FS • cos 03 • sin 203 - tan <(> • sin 03 • sin 203)
FS • (y • L • H • sin 03 • cos 03 • sin 203) - FS • (y • L • H • cos 03 • tan 8 • sin 203)
- FS • (Ca • L • cos 03 • sin 203) - FS • (y • L • H • sin203 • tan • sin 203)
+ (y • L • H • cos 03 • tan 8 + Ca • L) • (tan <|) • sin 03 • sin 203) = FS • (2 • C • H • cos 03 + y • H2 • tan (}))
FS2 • f j • y • L • H • sin 2 203 J - FS • (y • L • H • cos2G3 • tan 8 • sin 203+ Ca • L • cos 03 • sin 203
+ y • L • H • sin203 • tan (j> • sin 203 + 2 • C • H • cos 03 + y • H2 • tan )
+ (Y • L • H • cos 03 • tan 8 + Ca • L) • (tan <|> • sin 03 • sin 203) =0
A-24
-------
Appendix "B"
Derivation of Required Tensile Strength of Geogrid
or Geotextile Reinforcement of Cover Soil on a Geomembrane
Reinforcement
(Geogrid or
Geotextile)
Ep-f T = EA =>T = EA-EP
Y' L • H • sin (03 - 5n)
E_ u
A ~
A
COS <))
cos SD
•\C-
D LFS
H yH
sin 03 sin 2l
2 1
- • tan <{)DJ
cos (())D
= 1,5D=8,
Y' L' H • sin (G3 - 5)
5
cos
A fc-H y-H2
6- -r-rr+ . ^^ •
L sin GJ sin 203
tan
cos (<)) + i
A-25
-------
Appendix "C"
Derivation of Geomembrane Tensile Stress
Due to Unbalanced Friction Forces
__ Resisting Force T + TL • W • L
H* J ^^ ^"•""•^•""•^•™™™^™""™«""™"«« ^^ BMW^HHB^^H^MIIM^miMIBBBM
Driving Force TU • W • L
(a) If % = TL, FS > 1, pure shear @ T u = TL
(b) If % < TL, FS > 1 , pure shear = T y, rest of (TL - TU) is not mobilized.
(c) If % > TL, FS may be >, = or < 1, which depend on the T value
whenFS = l =>TU-W-L
T = TU- W- L-TL- W-L
T = C + an- tan <(>, CTn = y • H • cos GJ
TU = CaU + Y ' H • cos 13 • tan 8u
TL = CgL + Y • H • cos 03 • tan 8L
T
H- cosGJ- (tan8u -
tc_ gal'ow
ro ^ ^^"^^^^~
^reqd
where aallow =
-•uk
(FS)T
auit= Geomembrane Wide Width Tensile Test
A-26
-------
Appendix "D"
Derivation of Geomembrane Tensile Stress Due to
Subsidence of Material Beneath the Geomembrane
Cover Soil
Unit Weight'Ycs"
D
Geomembrane
C (circumference) = 2- K • L
t = thickness of the geomembrane
= (R-D)
R2 = R2 - 2 • R • D + D2 + L2
R =
D2 + L2
2-D
R
R-D
CL
(2-TC • r) • dr • Ycs>Hcs'r='
Jo
C • R
= arcqd- r (2-7C-L)- R
r\
•j • n • L • YCS • Hcs
°rc£id= t • (2 • K • L) • R
1? • Ya? ' HCS
3- t-R
2 • D • L2 • YCS ' HCS
3-f(D2 + L2)
A-27
-------
APPENDIX B
LONG-TERM DURABILITY AND AGING OF GEOMEMBRANES
-------
Long-Term Durability and Aging of Geomembranes
Robert M. Koerner1 , Yick H. Halse' and
Arthur E. Lord, Jr.'
Perhaps the most frequently asked question regarding
geomembranes (or any other type of geosynthetic material)
is, "how long will they last"? The answer to this
question is illusive (in spite of a relatively large data
base on polymer degradation) mainly because of the buried
nature of geomembranes. Soil burial greatly diminishes,
and even eliminates many of the degradation processes and
synergistic effects which have been most widely
investigated by the polymer industry for exposed
plastics. However, different degradation processes coming
from chemical Interactions and extremely long time frames
may be involved via exposure to liquids like leachate for
systems intended to last for many decades or even
hundreds of years. Thus the lifetimes of buried
geomembranes can be significantly different than exposed
plastics, but a quantitative method to predict "how long"
is still not available.
This paper describes the various degradation
mechanisms of plastics on an individual basis and then
addresses the various synergistic effects which may
accelerate degradation. It will be seen that synergistic
effects greatly complicate the situation. While
accelerated test methods are attractive to assess the
various phenomena, these procedures may significantly
misrepresent the actual long-term performance of
geomembranes. Thus the transfer of information must
proceed with caution.
'BowmanProfessor of Civil Engineering and Director,
Geosynthetic Research Institute, Drexel University
Philadelphia, PA, 19104.
'Research Assistant Professor, Geosynthetic Research
Institute, Drexel University, Philadelphia, PA, 19104. ,
'Professor of Physics, Geosynthetic Research Institute,
Drexel University, Philadelphia, PA, 19104.
106
GEOMEMBRANES DURABILITY AND AGING
GRI-96
The paper summarizes the different geomembrane
degradation processes and then concludes with some
suggested procedures on long-term simulation testing
coupled.with observations on field behavior.
Overview of Ggomembranes
Beginning with swimming pool liners (Staff 1984) made
from PVC sheeting in the 1930's, and extending to
bituminuous panels (Geier and Morrison 1968) for canal
liners, the geomembrane era began in earnest with
reservoir liners made from butyl rubber (Chuck 1970) in
the 1950's. These thermoset elastomers were made from
synthetic rubber which was developed during World War II.
To date, some manufacturers still refer to geomembranes
as "pond liners'". Other names have also arisen; for
example, flexible membrane liners (FML's), a term used
extensively by the U.S. EPA, synthetic membrane liners
(SML's), membrane liners, or, simply, liners. Being a
subset of the geosynthetics area, we will refer to them
as geomembranes. ASTM defines geomembranes as "an
essentially impermeable membrane used with foundation,
soil, rock, earth or any other geotechnical engineering
related material as an integral part of a man-made
project, structure or system".
Throughout the 1960's and the 1970's a tremendous
variety of geomembrane formulations were developed. They
were ao plentiful that they almost defied classification.
Many were blends of polymers and even blends of polymers
and nonpolymers. Copolymers were developed which gave
additional options. Even the manufacturing process gave
rise to many variations. The EPA Technical Guidance
Document (Matrecon, Inc. 1988) and Kay's book (1988) gives
some insight into the varieties of geomembranes available
during this period. The major use of geomembranes,
however, did not arrive until 1982 when the U.S. EPA
required that a liner must "prevent" pollution migration,
rather than only "minimize" pollution migration (U.S.
Federal Register 1982). This legislation essentially
required the use of FML's, or geomembranes, as being
preferred over clay liners.
Regarding the culling out of the large variety of
geomembrane types existing at that time, another EPA
document, this time in the form of a test protocol, was
significant. In 1984, EPA decided that a test method for
chemical compatibility, or resistance, was necessary so
that all EPA Regions and State Environmental Departments
would be permitting candidate geomembranes against the
same test standard. This test method, known as EPA Method
9090 (U.S. EPA 1984), has had the effect of greatly
reducing the many possible types of geomembranes in
-------
Long-Term Durability and Aging of Geomembranes
and
Robert M. Koerner1 , yick H. Halse1
Arthur E. Lord, Jr.'
Abstract
Perhaps the most frequently asked question regarding
geomembranes (or any other type of geosynthetlc material)
is, "how long will they last"? The answer to this
question is illusive (in spite of a relatively large data
base on polymer degradation) mainly because of the buried
nature of geomembranes. Soil burial greatly diminishes,
and even eliminates many of the degradation processes and
synergistic effects which have been most widely
investigated by the polymer industry for exposed
plastics. However, different degradation processes coming
00 from chemical interactions and extremely long time frames
i may be involved via exposure to liquids like leachate for
systems intended to last for many decades or even
hundreds of years. Thus the lifetimes of buried
geomembranes can be significantly different than exposed
plastics, but a quantitative method to predict "how long"
is still not available.
This paper describes the various degradation
mechanisms of plastics on an individual basis and then
addresses the various synergistic effects which may
accelerate degradation. It will be seen that synergistic
effects greatly complicate the situation. While
accelerated test methods are attractive to assess the
various phenomena, these procedures may significantly
misrepresent the actual long-term performance of
geomembranes. Thus the transfer of information must
proceed with caution.
'Bowman Professor of Civil Engineering and Director,
Geosynthetic Research Institute, Drexel University
Philadelphia, PA, 19104.
•Research Assistant Professor, Geosynthetic Research
Institute, Drexel University, Philadelphia, PA, 19104.
'Professor of Physics, Geosynthetic Research Institute,
Drexel University, Philadelphia, PA, 19104.
106
GEOMEMBRANES DURABILITY AND AGING
107
The paper summarizes the different geomembrane
degradation processes and then concludes with some
suggested procedures on long-term simulation testing
coupled.with observations on field behavior.
Overview of Geomembranes
Beginning with swimming pool liners (Staff 1984) made
from PVC sheeting in the 1930's, and extending to
bituminuous panels (Geier and Morrison 1968) for canal
liners, the geomembrane era began in earnest with
reservoir liners made from butyl rubber (Chuck 1970) in
the 1950's. These thermoset elastomers were made from
synthetic rubber which was developed during World War II.
To date, some manufacturers still refer to geomembranes
as "pond liners". Other names have also arisen; for
example, flexible membrane liners (FML's), a term used
extensively by the U.S. EPA, synthetic membrane liners
(SML1s), membrane liners, or, simply, liners. Being a
subset of the geosynthetics area, we will refer to them
as geomembranes. ASTM defines geomembranes as "an
essentially impermeable membrane used with foundation,
soil, rock, earth or any other geotechnical engineering
related material as an integral part of a man-made
project, structure or system".
Throughout the 1960's and the 1970's a tremendous
variety of geomembrane formulations were developed. They
were so plentiful that they almost defied classification.
Many were blends of polymers and even blends of polymers
and nonpolymers. Copolymers were developed which gave
additional options. Even the manufacturing process gave
rise to many variations. The EPA Technical Guidance
Document (Matrecon, Inc. 1988) and Kay's book (1988) gives
some insight into the varieties of geomembranes available
during this period. The major use of geomembranes,
however, did not arrive until 1982 when the U.S. EPA
required that a liner must "prevent" pollution migration,
rather than only "minimize" pollution migration (U.S.
Federal Register 1982). This legislation essentially
required the use of FML's, or geomembranes, as being
preferred over clay liners.
Regarding the culling out of the large variety of
geomembrane types existing at that time, another EPA
document, this time in the form of a test protocol, was
significant. In 1984, EPA decided that a test method for
chemical compatibility, or resistance, was necessary so
that all EPA Regions and State Environmental Departments
would be permitting candidate geomembranes against the
same test standard. This test method, known as EPA Method
9090 (U.S. EPA 1984), has had the effect of greatly
reducing the many possible types of geomembranes in
-------
108
WASTE CONTAINMENT SYSTEMS
current use, particularly those used for pollution
control. The test method has a parallel document directed
at assessing the test results, which is in the form of an
expert computer system called FLEX (U.S. EPA 1987). In
the next section, characterization of the most widely
used geomembranes is presented.
Types and Properties of Commonly Used Geomembranes
With the initial caution that polymer formulation and
processing is an on-going and constantly changing
technology, we will attempt to categorize the
geomembranes that are in common use. To be noted,
however, is that our perspective is from a geosynthetic
engineering design point of view. A different grouping
would indeed occur from a polymer chemist's or compound
formulator's perspective.
(a) Stiff (Semi-Crystalline) Thermoplastic Geomembranes -
In the stiff, semi-crystalline, thermoplastic category
are those geomembranes with crystallinity near, or above,
50% which results in stiffness values of greater than
1000 g-cm as per the ASTM D-1388 flexural rigidity test.
This is a bending test developed for fabrics, but it can
be readily used to distinguish stiff-from-flexible
geomembranes.
By far the most important geomembrane in this
category is high density polyethylene (HOPE). There are a
number of variations within HOPE, an important one being
the development of textured sheet which results in a high
friction surface. Also coextrusion manufacturing can
result in very intriguing composite materials with the
basic sheet being HDPE.
As a polymer formulation, HDPE is almost pure
polyethylene resin (about 97%), in the 0.935 to 0.937
g/cc density range. When carbon black is added for
ultraviolet stability, however, the material's overall
density is at, or slightly above, the 0.941 g/cc lower
limit of ASTM definition of high density polyethylene.
Thus it is commonly referred to as HDPE. The balance of
the compound is 2.0% to 2.5% carbon black and the
remaining 1.0% to 0.5% is antioxidant and processing
aide. The processing aides provide viscosity control,
manufacturing lubrication, and prevent adhesion between
various surfaces. They have also been called viscosity
depressants, slip agents and antiblocking agents. They
are generally proprietary as to their exact chemical
compositions. It should also be noted that the
polyethylene resins themselves have certain uniquenesses,
particularly the nature and extent of tie molecule
bonding and branching.
GEOMEMBRANES DURABILITY AND AGING
109
(b) Flexible (Low Crvsta1 Unity 1 Thermoplastic
Geomembranes - This group of geomembranes is
characterized by stiffness values in the ASTM D-1388
flexunel rigidity test as being significantly lower than
1000 g-cm. The main geomembranes in this category are
PVC, CPE, CSPE and VLDPE. Some typical values of flexural
rigidity are the following:
• 0.50 mm polyvinyl chloride (PVC) = 4 g-cm
• 0.75 mm chlorinated polyethylene (CPE) = 20 g-cm
• 0.75 mm chlorosulfonated polyethylene (CSPE) = 25 g-cm
• 0.75 mm very low density polyethylene (VLDPE) = 78 g-cm
While all of the above geomembranes have some
crystallinity, it is very low in comparison to HDPE. Yet,
all are indeed thermoplastic polymers and can be seamed
using thermal methods. The formulations of these
materials vary widely. Some typical values follow in
Table 1.
Table 1 - Typical Formulations for Flexible,
Thermoplastic, Geomembranes
Geo-
membrane
PVC
CPE
CSPE
VLDPE**
Resin
(%)
45-50
60-75
45-50
96-98
Plasticizer
(%)
35-40
10-15
2-5
0
Carbon Black
& Filler
(%)
10-15
20-30
45-50
2-3
Additive*
3-5
3-5
2-4
1-2
*refers to antioxidant, processing aids and lubricants
**note that this formulation is typical of most
polyethylenes, including HDPE
(c)Beinii
d. Flexible (Low Crystallinity1 Thermoplastic
Keomemranps - The behavior of geomembranes when they are
fabric reinforced (either via an internal scrim or via
spread coating) is very different than the unreinforced
variety of the exact same polymer. Certainly the short-
term engineering design properties are greatly altered
and conceivably the long-term properties may be altered
as well. The geomembranes are indeed flexible, however,
as the following data indicates. The stiffness values
-------
110
WASTE CONTAINMENT SYSTEMS
indicated are via the ASTM D-1388 test ,for flexural
rigidity:
• 0.91 mm chlorinated polyethylene (CPE-R)- 25 g-cm
• 0.91 mm chlorosulfonated polyethylene (CSPE-R)- 30 g-cm
• 0.66 mm ethylene interpolymer alloy (EIA-R)- 60 g-cm
As with the unreinforced materials just mentioned, all
can be seamed using thermal methods and are truly
thermoplastic. The formulations of the polymer component
are the same as given in Table 1.
Regarding the type of fabric used for the scrim
reinforcement of CPE-R and CSPE-R there are many
possibilities. Woven fabrics of high tenacity polyester
or nylon are the most common. They are often in a pattern
of 10 yarns per inch (4 yarns per centimeter) in both
directions, which is called a 10 x 10 scrim. Other
variations are 6x6 and 20 x 20 patterns, as well as
unbalanced variations. For reinforced geomembranes like
EIA-R, a very tightly woven fabric is used. Not
specifically mentioned are spread coated fabrics where a
variety of nonwoven needlepunched fabrics can be
utilized.
(d) Other Geomembranea - While the focus in this paper
_. will be on the geomembranes just reviewed, there are many
i other possibilities. Two complete groups have been
omitted since they are not currently used to any
significant degree in North America. They are the
thermoset elastomers (Matrecon 1988, Kays 1988) and
polymer modified bitumens (Gamski 1984), the latter group
being used quite often in Europe. Other newer materials
which fall within the categories just mentioned are in
the development stage and undoubtedly more will appear in
the future. This review, however, will be sufficient to
build upon in describing the various degradation
processes and synergistic effects which may occur.
Mechanisms of Degradation
This section of the paper describes various polymer
degradation processes which can act within a geomembrane.
Each process and its resulting implication is taken by
itself as though it were acting in isolation. This, of
course, is not field representative, but we feel that it
is necessary to describe the isolated events before
synergistic effects can be considered. Please note at the
outset that our perspective is from a gee-synthetic
engineering design point of view. From a chemistry or
polymeric science point of view there are many references
treating the various subjects in a much more rigorous
(and undoubtedly enlightened) manner.
GEOMEMBRANES DURABILITY AND AGING
111
(a) Ultraviolet Degradation - As shown in Figure 1, the
spectrum of natural light is broken into two major
regions (visible and ultraviolet) according to the
wavelength of the solar radiation. It is well established
in the! polymer literature that certain wavelengths within
the ultraviolet portion are particularly degrading to
polymeric materials. Van Zaten (1986) makes mention of
the following commonly used polymers and their most
sensitive wavelengths, all of which are in the
ultraviolet region and are noted on Figure 1.
• polyethylene = 300 nm
• polyester = 325 nm
• polypropylene = 3"70 nm
Furthermore, the mechanism of degradation is also well
understood. The light with the most sensitive wavelengths
enters into the molecular structure of the polymer
liberating free radicals which cause bond scission in the
primary bonding of the polymer's backbone. This
mechanism, in direct proportion to the intensity, causes
a reduction in mechanical properties to the eventual
point where the polymer becomes brittle and cracks to
unacceptable levels.
The above type of degradation is greatly reduced by
the use of carbon black or chemical based light
stabilizers. Carbon black is a finely dispersed powder
of approximately micron size which acts as a blocking (or
screening) agent to prevent the ultraviolet light from
entering into the polymer structure. It also absorbs some
of the energy. Its effectiveness decreases uniformly with
time of exposure so that the amount and dispersion of the
carbon black is important (Apse 1989). The maximum amount,
however, is limited to the amount which interferes with
the growth and strength of the polymer structure.
Hindered amine light stabilizers (HALS) are chemicals
added to the polymer compound which react with the free
radicals liberated by the ultraviolet light preventing
the propagation of degradation. When such additives are
consumed, however, continued ultraviolet exposure will
cause rapid degradation of the polymer. A combination of
carbon black and chemical absorbers has been shown to be
very effective in avoiding ultraviolet induced
degradation of polymers (Grassie and Scott 1985).
For geomembrane applications, a soil backfill or
other covering eliminates the problem of ultraviolet
degradation entirely. Only exposed geomembranes are
subjected to ultraviolet degradation and as little as 15
cm of soil cover is sufficient to prevent its occurrence.
Obviously, this cover soil must be placed in a timely
fashion which can be achieved in all except the following
-------
114
WASTE CONTAINMENT SYSTEMS
polyethylene and polypropylene geomembranes were
S?S neC*tet by the fadiation- Furthermore, the radiation
ni« «H ^ * * si*nificant effect on other chemical
degradation rates.
With the absence of radioactive materials in the
waste material, however, the subject becomes a moot
point Fortunately, this is the case for most solid and
liquid waste contained by geomembranes and related
geosynthetic materials.
(c) Chemical DearatlfiMnn - The reaction of various
geomembranes to chemicals has probably been studied more
than any other liner degradation mechanism. Most of the
work is laboratory oriented via simple immersion tests
but the body of knowledge is so great that a reasonable
confidence level can be associated with manufacturers
listings and recommendations. Complex waste streams like
leachate, however, are usually not addressed and must be
evaluated on a site specific basis. For this reason the
U.S. EPA developed the Method 9090 procedure. Here
samples of the candidate geomembrane are exposed at 23°C
and at 50°C and removed at 30, 60, 90 and 120 days
Various physical and mechanical tests are performed and
then compared to the unexposed geomembrane. A percent
change in this behavior is calculated. When plotted for
W the various exposure times, trends can be established and
<-n a decision made as to the nature and degree of chemical
resistance.
Depending on the type of leachate vis-a-vis the
polymeric compound from which the geomembrane is made, a
number of reactions may occur.'
•No reaction may occur, which indicates that the
geomembrane is resistant to the leachate; at least for
the time periods and temperatures evaluated.
• Swelling of the geomembrane may occur which in itself
may not be significant. Many polymers can accommodate
liquid in their amorphous regions without a sacrifice of
physical or mechanical properties. Swelling, however, is
often the first stage of subsequent degradation and a
small loss in modulus and strength may occur. The effect
is often reversible when the liquid is removed.
• Change of physical and mechanical properties, of course,
signifies some type of chemical reaction. The variations
are enormous. Quite often the elongation at break in a
tensile test will be the first property to show signs of
change. It will first occur with the 50°C incubation
data, since this can be considered to be an accelerated
test over the 23°C incubation data.
• A large change of physical and mechanical properties
signifies an unacceptable performance of the
GEOMEMBRANES DURABILITY AND AGING
115
geomembrane. Limits of acceptability are, however, very
subjective. Table 2 gives a number of recommendations as
accumulated by Koerner, 1990. It should be noted that
there is also an expert computer code available to aid
in t^e decision.
Table 2 - Suggested Limits of Different Test Values for
Incubated Geomembranes, (see the Reference by
Koerner, 1990, for other details and complete
references)
0.9 g/m2/hr
<20
<30
<30
10 points
>20
>30
>30
10 points
(b) For Stiff, Semi-Crystalline, Thermoplastic, Polymers
O'Toole
Little
Koerner
Property
Resis- Not Resis- Not Resis- Not
tant Resis- tant Resis- tant Resis-
tant tant tant
Permeation Rate
(water vapor)
(g/m2-hr) - - <0.9 20.9
Change in Weight
(%) <0.5 >1.0 <3 23
Change in Volume
(%) <0.2 >0.5 <1 21
Change in Yield
strength (%) <10 >20 <20 220
Change in Yield
Elongation (%) - - <20 220
Change in Modulus
(%) -
Change in Tear
Strength (%) -
Change in Puncture
Strength (%) -
<0.9
<2
<1
<20
<30
<30
<20
<30
20.9
22
21
220
230
230
220
230
-------
112
WASTE CONTAINMENT SYSTEMS
270 290 310 330 350 370 390 410 430 450 470 490 510 530 550 570 590
fPEjPET jpp W«vtl«ngth - nanonwitrt
Flgurt 1 - The W«v«hngth Spectrum ol Vltlbl* and UV Solir Reflation.
00
100
SnputolofiolDISM
HOOT FilimPoMi
<1.000 AlL*ol3
10-1.000 MUMI3
1.000-10.000 MLM*I
Mwe.000 MLM*3
Atar 10.000 M LM* 1
1 10 100 1000 10000 100000 1000000
Failure Time (hrs.)
Figure 2 - Schematic Plot of Time of Failure versus Pipe Hoop Stress
for Burst Testing of Un-Noicfied PE Pipe.
GEOMEMBRANES DURABILITY AND AGING
113
situations.
• Surface impoundments above the liquid level and along
theif horizontal runout length
•Canal liners above the liquid level and along their
horizontal runout length
•Covers of surface impoundments, i.e., floating covers
•Landfill liners on side slopes which have had their
surfaces exposed by erosion of cover soil and are
inaccessible.
For the above situations of exposed geomembranes some
amount of degradation over time is unavoidable.
(b) p*H
-------
116
WASTE CONTAINMENT SYSTEMS
(d) Degradation by Swelling - One indication of a
geomembrane's durability is the amount of swelling that
occurs due to liquid absorption. It should be emphasized
that swelling per se does not necessarily mean chain
scission nor a failed system. It is, however, a bit
disconcerting, and usually results in a change of
physical and mechanical properties, at least on a
temporary basis.
The test for water absorption, which can be modified
for any liquid, is given in ASTM D570. The test is
directed at a quantitative determination of the amount of
water absorbed, but it is also used as a quality control
test on the uniformity of the finished product. The test
procedure cautions that the liquid absorption may be
significantly different through the edge or through the
surface, particularly with laminated products. (This fact
alone suggests that in seaming of laminated geomembranes,
the upper overlap must be protected against moisture
uptake.) Test specimens of 75 by 25 mm are used and
immersed in a number of possible ways:
• For 2 hr., 24 hr., or 2 weeks of constant immersion in
23°C water.
• Under cyclic (repeated) immersion.
7° • For 0.5 hr. or 2 hr. of constant immersion in 50°C
^ water.
•For 0.5 hr. or 2 hr. of constant immersion in boiling
water.
The resulting test data are reported as the percentage
increase in weight using deionized and distilled water.
Some typical values for commonly used geomembranes are as
follows (Haxo, Nelson and Miedema 1985):
PVC - 3 to 4%
CPE - 1 to 4%
CSPE - 1 to 5%
HOPE - negligible
VLDPE — not evaluated
Swelling due to other liquids is mentioned in the
reference cited.
(e) Degradation hy Extraction - Some polymers exhibit
degradation by the long-term extraction of one or more
components of the compound from the polymeric material.
These are usually polymers which have been compounded
with the use of plasticizers and/or fillers. The as-
formulated and compounded mixture of such polymers is
very intricate and the bonding mechanism is very complex.
When extraction of plasticizers does occur, a sticky
surface on the geomembrane results with the remaining
GEOMEMBRANES DURABILITY AND AGING
117
structure showing signs of increased modulus and
strength, and a lowering of the elongation at failure,
i.e. the material becomes progressively brittle (Doyle
and Baker 1989) . The long term behavior, however, is
unknown. It is also possible that anti-degradient
components within the polymer may be extracted and leach
out to the surface. This might indicate that the
remaining polymer is somewhat more sensitive to long-term
degradation.
(f) Degradation by Delamination - For geomembranes which
are manufactured by calendering or spread coating,
delamination is a possibility. It is usually observed
when liquid enters into the edge of the geomembrane and
is drawn into the interface by capillary tension. This
can occur between geomembrane plies, between reinforcing
scrim and one of the plies, or between the geomembrane
coating of a fabric substrate as in spread coated
geomembranes. When it occurs, the individual components
are separated and composite action is lost. This type of
wicking action has been problematic in the past but
current manufacturing methods and proper CQC/CQA in field
operations have almost eliminated the situation.
(g) Oxidation Degradation - Whenever a free radical is
created, e.g., on a carbon atom in the polyethylene
chain, oxygen can create large scale degradation. The
oxygen combines with the free radical to form a
hydroperoxy radical, which is passed around within the
molecular structure. It eventually reacts with another
polymer chain creating a new free radical causing chain
scission. The reaction generally accelerates once it is
triggered as shown in the following equations.
R' + O -> ROO
ROO + RH -» ROOH
where
R" - free radical
ROO" - hydroperoxy free radical
RH - polymer chain
ROOH - oxidized polymer chain
Anti-oxidation additives are added to the compound to
scavenge these free radicals in order to halt, or at
least to interfere with, the process. These additives, or
stabilizers, are specific to each type of resin. This
area is very sophisticated and quite advanced with all
resin manufacturers being involved in a meaningful and
positive way. The specific anti-oxidants are usually
proprietary. Removal of oxygen from the geomembrane's
-------
118
WASTE CONTAINMENT SYSTEMS
surface, of course, eliminates the concern. Thus once
placed and covered with waste, or liquid, degradation by
oxidation should be greatly retarded. Conversely, exposed
geomembranes, or those covered by nonsaturated soil, will
be susceptible to the phenomenon.
-------
120
WASTE CONTAINMENT SYSTEMS
degradation and aging although the coupons are rarely, if
ever, in a stressed condition.
Accelerated Testing Methods
Clearly, the long time frames involved in evaluating
individual degradation mechanisms at field related
temperatures, compounded by synergistic effects, are not
providing answers regarding geomembrane behavior fast
enough for the decision making processes of today. Thus
accelerated testing, either by high stress, elevated
temperatures and/or aggressive liquids, is very
compelling. Before reviewing these procedures, however,
it must be clearly recognized that one is assuming that
the high stress, elevated temperature or aggressive
liquids used actually simulates extended lifetimes... an
assumption which is not readily substantiated. Thus it
might be that the test procedures to be described here
actually form lower-bound conclusions in predicting
degradation, i.e., the results may be minimum values but
that is not known with any degree of certainty.
(a) Stress Limit Testing - Focusing almost exclusively
on HOPE pipe for natural gas transmission, the Gas
Research Institute, the Plastic Pipe Institute and the
American Gas Association are all very active in various
aspects of plastic pipe research and development. The
three above-mentioned organizations, together with
Battelle Columbus Laboratories sponsor the Plastic Fuel
Gas Symposia which are held on a biennial basis and the
resulting Proceedings contain many interesting papers.
Stress limit testing in the plastic pipe area has
proceeded to a point where there are generally accepted
testing methods and standards. ASTM D1598 describes a
standard experimental procedure and ASTM D2837 gives
guidance on the interpretation of the results of the
D1598 test method.
In ASTM D1598, long pieces of un-notched pipe are
tightly capped and placed in a constant temperature
environment. Room temperature of 23°C is usually used.
The pipes are placed under various internal pressures
which mobilize different values of hoop stress in the
pipe walls, and the pipes are monitored until failure
occurs. This is indicated by a sudden loss of pressure.
Then the values of hoop stress are plotted versus failure
times on a log-log scale, see Figure 2. If the plot is
reasonably linear, a straight line is extrapolated to the
desired, or design, lifetime which is often 105 hours or
11.4 years. The stress at this failure time multiplied by
an appropriate factor is called the hydrostatic design
basis stress. While of interest for pipelines, the stress
state of geomembranes is essentially unknown and is
CEOMEMBRANES DURABILITY AND AGING
121
extremely difficult to model. Thus the technique is not
of direct value for geomembrane design. It leads,
however, to the next method.
(b) pate Process Method (RPM1 for Pipes - Research at the
Gas Institute of The Netherlands (Wolters 1987) uses the
method of pipe aging that is most prevalent in Europe. It
is also an International Standards Organization (ISO)
tentative standard currently in committee. Note that
other plastic pipe research institutes also are involved
in this type of research. The experiments are again
performed using long pieces of un-notched pipe which are
tightly capped, but now they are placed in various
constant temperature environments. So as to accelerate
the process, elevated temperature baths up to 80°C are
used. Different pressures are put in the pipes at each
temperature so that hoop stress occurs in the pipe walls.
The pipes are monitored until failure occurs, resulting
in sudden loss of pressure. Two distinct types of
failures are found; ductile and brittle. The failure
times corresponding to each applied pressure are
recorded. A response curve is presented by plotting hoop
stress against failure time on a log-log scale.
The rate process method (RPM) is then used to predict
a failure curve at some temperature other than those
tested, i.e., at a lower (field related) temperature than
was evaluated in the high temperature tests. This method
is based on an absolute reaction rate theory as developed
by Tobolsky and Eyring (1943) for viscoelastic
phenomenon. Coleman (1956) has applied it to explain the
failure of polymeric fibers. The relationship between
failure time and stress is expressed in the form as
follows:
log t.
V
-1
-1
A T a
where
tf = time to failure
T - temperature
a - tensile stress on the fiber
A0 and A1 - constants
Bragaw (1983) has revised the above model on polymeric
fibers and found three additional equations which yield
reasonable correlation to the failure data of HOPE pipe.
These three equations are as follows:
-------
122
WASTE CONTAINMENT SYSTEMS
log tf - A + AT + AT P
log tf - AQ +
log tf - AQ +
A log P
log P
(2)
(3)
(4)
where
P - Internal pipe pressure proportional to the hoop
stress in the pipe
The application of RPM requires a minimum of two
experimental failure curves at different elevated
temperatures generally above 40°C. The equation which
yields the best correlation to these curves is then used
in the prediction procedure for a response curve at a
field related temperature e.g., 10°C to 258C. Two
separate extrapolations are required, one for the ductile
response and one for the brittle response. Three
representative points are chosen on the ductile regions
of the two experimental curves. One curve will be
selected for two points, and the other, the remaining
03 point. This data is substituted into the chosen equation,
^ i.e., Equation 1, 2 or 3, to obtain the prediction
° equation for the ductile response of the curve at the
desired (lower) temperature. The process is now repeated
for the predicted brittle response curve at the same
desired temperature. The intersection of these two lines
defines the transition time.
Figure 3 shows two experimental failure curves which
were conducted at temperatures of 80°C and 60°C along
with the predicted curve at 20°C. The intersection of the
linear portions of the 20°C curve represents the
anticipated time for transition in the HPDE pipe from a
ductile to a brittle behavior of the material. For pipe
design, however, the intersection of the desired service
lifetime, say 50 years, with the brittle curve is the
focal point. A factor of safety is then placed on this
value, e.g., note that it is lowered to 6.5 MPa in Figure
3. This value of stress is used as a limiting value for
internal pressure in the pipe.
(c) Rate. — Pr.QC.e33
(KPN) for Geomembranes - A
similar RPM method to that just described for HOPE pipes
can be applied to HOPE geomembranes . The major difference
is the method of stressing the material. The geomembrane
tests are performed using a notched constant load test
(NCLT) . In this test, dumbbell shaped specimens are taken
from the geomembrane sheet. A notch is introduced on one
of the surfaces; the notch depth being 20% of the
thickness of the sheet. The full description of the
GEOMEMBRANES DURABILITY AND AGING
123
logo
(Mpa)
"allow
= 6.5
log rime
50 yrs.
Figure 3 - Burst Te$ Data for MDPE Pipe. The Intersection of the Ductile Portion
of the 20 C Line and 50 Years Has Been Lowered to 6.5 MPa by
Multiplication with the Appropriate Factor of Safety.
(From Ref. 14)
100
CO
in
CD
55
33
O)
c
50 C
.1
10
Failure Time (hours)
100
1000
Figure 4 - Notched Constant Load Tests (NCLT) on HOPE Geomembrane
Samples Immersed in 10% lgepal/90% Tap Water Solution.
-------
124
WASTE CONTAINMENT SYSTEMS
notching process is described by Halse, et al. (1990).
Tensile loads varying from 30% to 70% of the yield stress
of the sheet, are applied to the notched specimens. The
tests are performed in elevated constant temperature
environments (usually from 40°C to 80°) and in a surface
active wetting agent. Often a 10% Igepal and 90% tap
water is used. The data is presented by plotting percent
yield stress against failure time on a log-log scale.
Figure 4 shows typical experimental curves at 50°C and
40°C which are seen to be very similar to the behavior of
HOPE pipe, recall Figure 3. Here distinct ductile and
brittle regions can be seen along with a clearly defined
transition time.
In order to use these elevated temperature curves to
obtain the transition time for a realistic temperature of
a geomembrane beneath solid waste or liquid impoundments,
e.g., at 25°C, only Equations 3 or 4 can be used due to
the data being plotted on a log-log scale. The following
is a step-by-step procedure of how the experimental data
is used in this regard.
Step. 1 Determine the best fit equation among Equations
3 or 4. We will select Equation 3 which has
been found to show the best fit for both the
ductile and brittle data.
Step. 2 Obtain the equation for the ductile region of
the 25°C curve. To predict the ductile region
of the 25°C curve, two data points were chosen
from the ductile region of the 50°C curve and
one from the 40°C curve. The details of their
values are as follows:
Point 1: T! - 323 °K; Pj - 48% = 9.72 MPa;
tx - 1 h
Point 2: T2 - 313 °K; P2 = 52% = 10.53 MPa;
t2 - 5 h
Point 3: T3 - 313 °K; P3 - 56% - 11.34 MPa;
t3 - 1 h
Note that the calculations are in degrees
Kelvin which is degrees Centigrade plus 273
degrees. By substituting the above three seta
of data into Equation (3), and solving the
resulting three simultaneous equations, three
constants are obtained and are listed below:
A - -37.18,
- 18774, and A2 - -21.2
By substituting these constants back into
GEOMEMBRANES DURABILITY AND AGING
125
Equation (3) the following equation results for
predicting the ductile portion of the curve at
any temperature.
log t = -87.18 + 18774/T - 21.2 log P
(5)
Step 3. Obtain the equation for the brittle region of
the 25°C curve. For the brittle region, two
data points were chosen from the brittle region
of the 50°C curve and one point from the 40°C
curve. The details of their values are as
follows :
Point 1:
= 323 °K;
37% = 7.50 MPa;
t1 = 10 h
Point 2: T2 = 323 °K; P2 = 28% =5.67 MPa;
t2 = 20 h
Point 3: T3 = 313 °K; P3 = 37% =7.50 MPa;
t3 - 30 h
The resulting three constants obtained by
simultaneous solution of the resulting three
forms of Equation 3 are listed below:
A0 = -11.8, A1 = 4853, and A2 = -2.5
Therefore, the equation for predicting the
brittle portion of the curve at any temperature
is as follows:
log t - -11.8 -»• 4853/T - 2.5 log P
(6)
Step 4. Use Equations (5) and (6) to obtain the
predicted response curves at the desired
temperature. The ductile behavior from Equation
(5) and the brittle behavior from Equation (6)
at the specific temperature of 25°C was used to
plot the desired curves at 25°C in Figure 4.
Step 5. Intersect the two prediction curves for the
ductile-to-brittle transition time. As seen in
Figure 4 this intersection produces a
transition time for the 25°C temperature
response of 60 hours. Note that the ductile-to-
brittle transition time is increased over ten
fold (from 6 to 70 hours) by decreasing 25
degrees in temperature. It should also be noted
that this curve represents the behavior of the
geomembrane in 10* Igepal solution. The
performance of the same geomembrane will
-------
126 WASTE CONTAINMENT SYSTEMS
probably be very different in other
environments, such as in leachate or water.
(d) Elevated Temperature and Arrhenius Modeling - Using
an experimental chamber as shown in Figure 5, Mitchell
and Spanner (1985) have superimposed compressive stress,
chemical exposure, elevated temperature and long testing
time into one experimental device. For their particular
tests, three duplicate chambers were operated at 18°C,
48°C and 78°C respectively. At the end of the arbitrarily
designated test period (in the example to be described it
was 18 weeks), the geomembrane samples were removed.
Mechanical tests and chemical analyses were performed on
these incubated samples to monitor if any changes in the
various properties of the geomembranes occurred.
The mechanical tests included the following:
• Tensile Strength and Elongation
• Yield Strength and Elongation
• Stress Cracking Behavior
to
The chemical analyses included the following:
• Differential scanning calorimetry (DSC); for
measuring crystallinity and oxidation induction
time (OIT)
• Infra-red spectrometry (IR); for measuring
concentration of carbonyl groups
• Gel permeation chromatography (GPC) ; for
measuring the molecular weight and molecular
weight distribution
If there were changes in any of the above properties; for
example, in the concentration of the carbonyl group, the
reaction rate (K) was obtained for each experimental test
temperature (T) . These values were now used with the
Arrhenius equation which is as follows (American National
Standard 1986) :
K
Ae
-E/RT
(4)
where
K
A
E
T
R
reaction rate for the process considered
a constant for the process considered
reaction activation energy for the process
considered
temperature (°K -°C + 273)
gas constant (8.314 J/mol-°K)
By plotting "In K" against "1/T", a straight line should
be obtained, see Figure 6. The slope of this line is
"-E/R" for the particular property change being
GEOMEMBRANES DURABILITY AND AGING
127
Hydraulic Load Device
Air (7)
Supply
Liquid Level
Sight Glass
*- — 'Record
1 — "Q
^~*.
c
I
1
0
c
0
c
o
c
o
er o
Liquid
1
^-«r
Sand
3
3
3
f
3
9
S'
Leachale
Recirculalion
Pump
-Heat Transfer
Coils
Perforated
Plate Press
-Thermocouples
-Test Liner
Drain
Figure 5 - Schematic of Accelerated Aging Column
(after Mitchell and Spanner)
CD
_D
0)
to
DC
O
•&
o
(0
a>
a:
In A
Governing Equation:
In K = In A - ()
78°C 48°C 18°C
Inverse Temperature (1/T)
Figure 6 - Graphical Method of Plotting Reaction Rate Values
for Change in a Specific Geomembrane Property.
-------
128
WASTE CONTAINMENT SYSTEMS
monitored. The constant "A" can also be identified but it
drops out of the equation when comparing the responses at
two different temperatures.
For example, Turi (1981) found that the activation
energy (E) of polyethylene is 109,000 J/mol. This is the
energy requirement for an atom to change its position. If
we use this value in Mitchell and Spanner's tests, then
the degradation that occurs between the two tests at
temperatures of 18°C and 78°C is determined as follows:
18 + 273
- e
(109.000) I 1
8.314 L351
29l
- 2254
Hence, for their particular process occurring in 18 weeks
at 78CC it would take 2264 times 18 weeks, or 784 years,
to occur at 18°C.
(e) Hoechst Multiparameter Approach - The Hoechst
research laboratory in West Germany has been active in
long term testing of HDPE pipe since the 1950's. Recently
they have been applying their expertise and experience to
the long term behavior of HDPE geomembranes (Kork, et al.
1987) . Note, however, that there are major differences
between pipe and sheet. The stress state in pressurized
pipe is well known, whereas the stress state in the
geomembranes in the field is not nearly as well known.
Furthermore, the pipe is under a constant stress state,
whereas stress relaxation can occur in geomembranes. If
these two differences in the state of stress between pipe
and sheet can be resolved, then the tremendous wealth of
data obtained in over 30 years of pipeline testing can be
carried over to the geomembrane area.
The Hoechst group has considered geomembranes for the
case of local subsidence under the liner and the
mobilization of out-of-plane deformation of the sheet
into an multi-axial stress state. Thus their model for
sheet stresses' Is biaxial stress. (In the usual
pressurized pipe, the radial stress is twice the value of
the longitudinal stress. Hence the isotropic biaxial
stress state in a normal pipe testing experiment was
achieved by putting the pipes under an additional
longitudinal stress. It was found that the lifetimes in
these tests were the same as in normal pressure burst
testing. Hence if biaxial stress relaxation could be
accounted for, a viable long term testing technique could
very well be developed for sheet from modified pipe
testing. This of course assumes that the different
manufacturing and processing of pipe and sheet produce
GEOMEMBRANES DURABILITY AND AGING
129
essentially the same material properties. (This may not
be completely the case.) The Hoechst long term testing
for geomembrane "sheet" thus consists of the following
procedure:
(1) Modified burst testing of pipe (of the same material
as the sheet), with additional longitudinal stress
to produce an isotropic, biaxial stress state.
Their tests make note that the site-specific liquid
should be used.
(2) Assume a given subsidence strain versus time
profile.
(3) Measurement of stress relaxation curves in sheets
which have been stressed biaxially, at strain values
encountered in field.
(4) Use steps (2) and (3) to predict the stress as a
function of time.
(5) See how these maximum stresses compare with the
stress-lifetime curves determined in the normal
constant stress-lifetime pipe measurements of Step
(1). The constant stress-1ifetime curves are
modified (as in the normal pipe testing) to
accommodate the effects of various chemicals and
even for seams.
(6) If linear accumulation of degradation is assumed,
then the variable stress curves can be used to
predict failure from the constant stress-1ifetime
curves.
The steps (l)-(5) have been performed by Kork, et al.
(1987) . The approach should certainly be considered
seriously. One of the main impediments to its viability
would seem to be that if one produces pipe, the final
product may have different material properties than
sheet. For example, the residual stresses could be quite
different. Other work of a related nature can be found in
Hessell and John (1987) and Gaube, et al. (1976).
Summary and Conclusions
This rather lengthy treatise on durability and aging
has attempted to give insight into long-term performance
of geomembranes by itemizing those mechanisms which can
degrade the polymer resin and/or compound. Table 3 gives
a summary of the individual degradation mechanisms that
were discussed. All of them, taken individually, are
possible to evaluate and/or quantify. In addition, some
suggestions as to preventative measures are offered.
-------
Table 3 - Degradation Phenomena In Geomembranes (from a GeosyntheUc Engineering Perspective)
Degradation Degradation
Classification Mechanism
Ultraviolet • chain
scission
• bond
breaking
Radiation • chain
scission
Chemical • reaction with
structure
• reaction with
additives
Swelling • liquid
absorption
Extraction • additive
expulsion
Initial Change In Material Subsequent Change In Material Preventative
Laboratory!" Field* Laboratory*31 Field*4' Measure
• mol. wL
•stress
crack
resist.
• moL wt.
•stress
crack
resist.
• carbonyl
index
• m
• mol. wt.
•TGA
•TGA
•IR
•color
• crazing
•color
•crazing
• texture
• color
• crazing
•TGA
• thickness
•color
•texture
•torture
•color
• thickness
• elongation
• modulus
• strength
• elongation
• modulus
• strength
• elongation
•modulus
• strength
•thickness
•modulus
• strength
• elongation
• modulus
• thickness
• color
• cracking
• color
• cracking
• texture
• cracking
• reactions
•thickness
•softness
• texture
• color
• screening
agent
• anti-
degradlent
•cover
geomembrane
• cover for fiand
a- rays
• shield for
neutrons
• reduce dosage
for 7 rays
• proper resin
•proper
additives
• proper resin
•proper
manufacturing
•proper
compounds
• proper
manufacturing
<
CO
- i
1
c»
Table 3 - Continued
Degradation
Classification
Delamlna-
tion
Oxidation
Biological
Degradation
Mechanism
• adhesion
loss
• reaction
with structure
• reactions
with additives
Initial Chance li
Laboratory*"
• thickness
• edge effects
• IR
• carbonyl
Index
• orr
• mol. wt.
• carbonyl
Index
• IR
[^Material Si
FleW«
• thickness
• edge effects
• color
• crazing
• texture
• surface
film
jhsequent Change In Material
Laboratory"31 Field*4'
• thickness
• ply
adhesion
• elongation
• modulus
• strength
• elongation
• modulus
• strength
• layer
separation
• thickness
• color
• cracking
• texture
• surface
layer
• cracking
Preventatlv*
Measure
• proper
manufacturing
• protect
geomembrane
edges
• antl-oxidant
• cover with
soil
• cover with
liquid
• avoid sensitive
additives
• bloclde
Notes:
: molecular weight. IR =
(1) Initial laboratory changes are generally sensed by chemical fingerprinting methods: mol. wt.
Infrared spectroscopy. TGA = thermal geometric analysis. OFT = oxidation inducatlon time
(2) Initial fleld changes are generally sensed on a qualitative basis.
(3) Subsequent laboratory changes can be sensed by numerous physical and mechanical tests. Listed In the table are those
considered to be the most sensitive parameters.
(4) Subsequent fleld changes are a continuation of the initial changes until physical and mechanical properties being to
visually change.
a
!
B-14
-------
J32
WASTE CONTAINMENT SYSTEMS
Also described in the paper were various synergistic
effects which greatly complicate the above mechanisms
when acting concurrently. Items such as elevated
temperature, biaxial or triaxial stress and long exposure
time are very difficult to model accurately.
Nevertheless, much laboratory modeling had been done,
although most has been by the plastic pipe industry. This
is almost exclusively on polyethylene pipe and the
technology transfer to polyethylene geomembranes is
certainly very valuable. Many of the techniques are being
evaluated for long-term geomembrane performance; for
example, ductile-to-brittle transition time along with
Arrhenius modeling. Other types of geomembranes appear to
be in need of long-term simulation testing as was
discussed in various facets of the paper.
In conclusion, it is felt that case histories (both
positive and negative) give the best insight into the
field behavior of geomembrane lined facilities at this
point in time. Test strips which are exhumed annually and
tested, versus the original material are extremely
valuable. They can be placed along the edge of the
facility or within sump areas for easy removal. Field
failures are also very important to analyze for aiding
and prompting in the modification of existing polymer
formulations and perhaps changing the basic resins
themselves. Lastly, long term laboratory tests under
simulated conditions should be undertaken. Simulated
stress tests under elevated temperature testing and
Arrhenius modeling is very intriguing in this regard as
is some type of modification of the Hoechst procedures.
Answers to the important question of "how long will they
last" may never be known unless the effort begins as soon
as possible.
----- , U. S. Federal Register, Regulations on Liner
Systems, July 26, 1982.
----- , U. S. EPA Method 9090, Compatibility Test for
Wastes and Membrane Liners, in Test Methods for
Evaluating Solid Waste Physical/Chemical Methods, SW-
846, 2nd Edition, 1984.
----- , U.S. EPA Computer Code on Flexible Membrane Liner
Advisory Expert System (FLEX), Cincinnati, Ohio, 1987.
« "Standard for Polymeric Materials - Long Term
Property Evaluations," American National Standard,
ANSI/UL 746B - 1986.
Apse, J. I., "Polyethylene Resins for Geomembrane
Applications," Durability and Aging of Cgnavothetir^,. R.
M. Koerner, Ed., Elsevier Appl. Sci. Publ. Ltd., 1989,
pp. 159-176.
GEOMEMBRANES DURABILITY AND AGING
133
Bragaw, G. G., "Service Rating of Polyethylene Piping
Systems by the Rate Process Method," 8th Plastic Fuel
Gas Pipe Symp., Nov. 1983.
Charlesby, A., Atomic Radiation and Polymers, Oxford
University Press, 1960.
Chuck, R. T., "Largest Butyl Rubber Lined Reservoir,"
Civil Engineering, ASCE, May 1970, pp. 44-47.
Coleman, B. D., "Application of the Theory of Absolute
Reaction Rates to the Creep Failure of Polymeric
Filaments," Jour. Polymer Science, Vol. 20, 1956, pp.
447-455.
Doyle, R. A. and Baker, K. C., "Weathering Tests of
Geomembranes," Durability and Aging of Geosynthetics. R.
M. Koerner, Ed., Elsevier Appl. Sci. Publ. Ltd., 1989,
pp. 152-158.
Gamski, K., "Geomembranes: Classification, Uses and
Performance," Jour. Geotex. and Geomemb., Elsevier Appl.
Sci. Publ. Ltd., Vol. 1, 1984, pp. 85-117.
Gaube, E., Diedrick, G. and Muller, W., "Pipes of
Thermoplastics; Experience of 20 Years of Pipe Testing,
Kunstoffe, Vol. 66, 1976, pp. 2-8.
Geier, F. H. and Morrison, W. R., "Buried Asphalt Membrane
Canal Lining," Research Report No. 12, A Water Resources
Technical Publication, Bureau of Reclamation, Denver,
Colorado, 1968.
Grassie, N. and Scott, G., Polymer Degradation and
Stabilization. Cambridge University Press, 1985.
Halse.Y. H., Lord, A. E. Jr. and Koerner, R. M., "Ductile-
to-Brittle Transition Time in Polyethylene Geomembrane
Sheet," Symposium on Geosynthetic Testing for Waste
Containment Applications, ASTM Spec. Tech. Publ., to
appear in 1990.
Haxo, H. E., Nelson, N. A. and Miederaa, T. A., "Solubility
Parameters for Predicting Membrane Waste Liquid
Compatibility," Proc. EPA Conf. on Hazardous Waste,
Cincinnati, OH, Apr. 1985, pp. 198-215.
Hessel, J. and John, P., "Long Term Strength of Welded
Joints in Polyethylene Sealing Sheets,"
Werkstofftechnik, Vol. 18, 1987, pp. 228-231.
Kays, W. B., Construction of Linings for Reservoirs. Tanks
and Pollution Control Facilities. 2nd Ed., J. Wiley and
Sons, Inc., NY, 1988.
Koerner, R. M., Designing with Geosvnthetics. 2nd Ed.,
Prentice Hall Publ. Co., Englewood Cliffs, NJ, 1990.
Kork, R., et. al., "Long Term Creep Resistance of Sheets
of Polyethylene Geomembrane, Report TR-88-0054 from
Hoechst A.G., Frankfurt, W. Germany, 1987.
Matrecon, Inc., Lining of Waste Impoundment and Disposal
Facilities, EPA/600/2-88.052, Sept., 1988.
Mitchell, D. H and Spanner, G. E., "Field Performance
Assessment of Synthetic Liners for Uranium Tailings
Ponds," Status Report, Battelle PNL, U.S. NRC, NUREG/CR-
4023, PNL-5005, Jan., 1985.
-------
*
c
I
o
o
3
a
134
WASTE CONTAINMENT SYSTEMS
Phillips, D. C., "Effects of Radiation on Polymers,"
Materials Science and Technology, Vol. 4, 1988, pp. 85-
91.
Staff, C. E., "The Foundation and Growth of the
Geomembrane Industry in the United States/" Intl. Conf.
Proc. on Geomembranes, Dever, Colorado, IFAI, 1984, pp.
5-8.
Tobolsky, A. and Eyring, H., "Mechanical Properties of
Polymeric Materials," Jour. Chem. Phys., Vol. 11, 1943,
pp. 125-134.
Turi, A., Thermal Characterization of Polymeric Materials.
Academic Press, 1981.
Van Zaten, R. V., Geotextiles and Geomembranes in Civil
Engineering. A. A. Balkema Press, Rotterdam and Boston,
1986.
Whyatt, G. A. and Farnsworth, R. K., "The Effect of
Radiation on the Properties of HOPE and PP Liners,"
Symposium on Geosynthetic Testing for Waste Containment
Applications, ASTM Spec. Tech. Publ., to appear in 1990.
Welters, M., "Prediction of Long-Term Strength of Plastic
Pipe," Proc. 10th Plastic Fuel Gas Symp., 1987, New
Orleans, Amer. Gas Assoc. Publ.
CO
Acknowledgements
The authors would like to express sincere
appreciation to all of the sponsoring member
organizations of the Geosynthetic Research Institute. A
listing of the firms and their contact members is
available from the authors. The Institute is focused on
long term generic research in geosynthetics of the type
described in this paper.
WASTE
CONTAINMENT
SYSTEMS:
Construction, Regulation, and Performance
Proceedings of a Symposium sponsored by the
Committee on Soil Improvement and Geosynthetics
and the Committee on Soil Properties of the
Geotechnical Engineering Division,
American Society of Civil Engineers
in conjunction with the
ASCE National Convention
San Francisco, California
November 6-7.1990
GEOTECHNICAL SPECIAL PUBLICATION NO. 26
Edited by Rudolph Bonaparte
Published by the
American Society of Civil Engineers
345 East 47th Street
New ftrk, New York 10017-2398
-------
United States
Environmental' Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268-1072
Official Business
Penalty for Private Use. $300
Please make all necessary changes on the above label,
detach or copy, and return to the address in the upper
left-hand corner
If you do not wish to receive these reports CHECK HERE D,
detach, or copy this cover, and return to the address in the
upper left-hand corner
EPA/625/4-91/025
------- |