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

-------