EPA/600/R-96/149
June 1996
REPORT OF 1995 WORKSHOP ON GEOSYNTHETIC CLAY LINERS
by
David E. Daniel and Heather B. Scranton
University of Texas at Austin
Department of Civil Engineering
Geotechnical Engineering Center
Austin, Texas 78712
Cooperative Agreement No. CR-821448-01-0
Project Officer
David A. Carson
U.S. Environmental Protection Agency
National Risk Management Research Laboratory
Cincinnati, Ohio 45268
NATIONAL RISK MANAGEMENT RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
Printed on Recycled Paper
-------
DISCLAIMER
The information in the document has been funded wholly or in part by the United States
Environmental Protection Agency under assistance agreement number CR-815546-01-0. It has
been subject to the Agency's peer and administrative review and has been approved for
publication as a U.S. EPA document. Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
-------
FOREWORD
The U-S. Environmental Protection Agency is charged by Congress with protecting the
Nation's land, air, and water resource. Under a mandate of national environmental laws, the
Agency strives to formulate and implement actions leading to a compatible balance between
human activities and the ability of natural systems to support and nurture life. To meet this
mandate, EPA's research program is providing data and technical support for solving
environmental problems today and building a science knowledge base necessary to manage our
ecological resources wisely, understand how pollutants affect our health, and prevent or reduce
environmental risks in the future.
The National Risk Management Research Laboratory is the Agency's center for
investigation of technological and management approaches for reducing risks from threats to
human health and the environment. The focus of the Laboratory's research program is on
methods for the prevention and control of pollution to air, land, water, and subsurface resources;
protection of water quality in public water systems; remediation of contaminated sites and
ground water; and prevention and control of indoor air pollution. The goal of this research effort
is to catalyze development and implementation of innovative, cost-effective environmental
technologies; develop scientific and engineering information needed by EPA to support
regulatory and policy decisions; and provide technical support and information transfer to ensure
effective implementation of environmental regulations and strategies.
This publication has been produced as part of the Laboratory's strategic long-term
research plan. It is published and made available by EPA's Office of Research and Development
to assist the user community and to link researchers with their clients.
E. Timothy Oppelt, Director
National Risk Management Research Laboratory
-------
ABSTRACT
A workshop was held at the EPA's National Risk Management Research Laboratory in
Cincinnati, Ohio, on August 9 and 10, 1995. On August 9, attendees where shown field plots of
GCLs that have been constructed at a site in Cincinnati, and given a detailed account of the test
plot layout, instrumentation, and performance to date. Fourteen test plots, with cross-sections
that are typical of landfill cover systems, were constructed. Five of the test plots were
constructed on 3H:1V slopes, and the remainder were constructed on 2H:1V slopes. All test
plots were two GCL roll widths wide (approximately 10 m), and 20 to 30 m long. The thickness
of the cover materials over the GCLs was approximately 900 mm. The objective of the project
was to verify that the GCLs would remain stable against mid-plane shear on 3H:1V slopes with a
factor of safety of at least 1.5 So long as the slopes are stable at 2H:1V, it can be demonstrated
that the minimum factor of safety is 1.5 for a 3H:1V slope. Thus far, all slopes have remained
stable with respect to mid-plane shear, although two interface failures did occur between the
GCL and an overlying textured geomembrane. Both failures occurred at the interface between
the woven geotextile component of a GCL and the geomembrane. Interface shear testing
performed after the failures demonstrated that the failures could have been predicted, based on
the laboratory shear test results. A significant finding from the project is that in landfill cover
applications, the internal strength of GCLs may not be the critical strength — interface strengths
may be lower than the internal strength of the GCL. Another significant finding has been that
bentonite encased between two geomembranes on one test plot has undergone unanticipated
hydration, when the expectation was that it would remain dry. The cause of hydration has not
been isolated but may be related to water migration through the penetrations made for
instrumentation cables.
On August 10, a series of presentations summarized recent research findings for the field
test site, as well as other on-going research projects. Significant information was presented
concerning laboratory shear testing results and seepage rates through composite
geomembrane/GCL liners based on flow rates in the leakage detection layer of double composite
liner systems.
This workshop marks the third in a series of workshops. The first two were held in 1990
and in 1992. The breadth and depth of information presented at the 1995 workshop is much
greater than in the earlier two workshops, and reflects the maturing of the GCL industry and the
efforts of many individuals to develop the technical information that is needed to evaluate and
assess GCLs as they are used in waste containment facilities.
IV
-------
Table of Contents
Page No.
Disclaimer ii
Foreword iii
Abstract iv
List of Figures ix
List of Tables xi
Acknowledgments xii
Chapter 1 — Introduction 1
Chapter! — Background on GCLs 3
2.1 Types of GCLs 3
2.2 Advantages of GCLs 5
2.3 Shear Strength of GCLs 6
2.3.1 Magnitude of Normal Stress 6
2.3.2 Water Content 7
2.3.3 Rate of Loading 7
2.3.4 Reinforcement 7
2.3.5 Amount of Deformation 8
2.3.6 Seismic Loading 9
2.4 Interface Shear Strength 9
Chapter 3 — Field Test Pbts in Cincinnati, Ohio 10
3.1 Expectations Concerning Field Performance at the Beginning of the Project n
3.2 Layout of the Test Plots 11
3.3 Plot Compositions 12
3.4 Anchor Trenches 17
3.5 Toe Detail ; 21
3.6 Instrumentation 23
*«O
f-
3.6.1 Moisture Sensors 24
3.6.2 Deformation Gauges 28
3.7 Construction 32
3.8 Cutting ope Geosynthetics 32
3.9 Field Peifitnance 33
3.9.1 Reformation Data 34
:3,9.\.l Total Displacement Data 34
fe.9.1.2 Relative Displacement 35
*3.9.1.3 Displacements after Materials in the Anchor 35
I Trench were Cut
§.9.1.4 Moisture Gauge Readings 35
6.9.1.5 Summary of Moisture and Deformation Data 35
- for all Plots
-------
3.9.2 Interface Slides
37
3.10 Plot F (Gundseal® with the Bentonite Side Facing Upward) 37
3.11 Additional Test Plot P (Gundseal® with the Bentonite Side 40
Facing Upward)
3.12 Erosion Control Materials 40
3.13 Summary/Future Plans 40
Chapter 4 — Research on GCLs 42
4.1 Shear Behavior of GCL Interfaces at the EPA Test Plots 42
in Cincinnati, Ohio
4.2 Aspects of GCL Performance of Composite Liners Containing GCLs 45
4.2.1 Field Hydraulic Performance of Composite Liners 45
Containing GCLs
4.2.2 Hydration of GCLs Adjacent to Subgrade Soil Layers 47
4.2.3 Causes of Failure of a Landfill Cover System Containing a GCL 48
4.2.4 Shear Strength of Hydrated GCLs at High Normal Stress 49
4.2.5 Shear Strength of GCL-Geomembrane Interfaces 50
4.3 University of Texas Research 50
4.3.1 Direct Shear Tests of GCL Bentonite 50
4.3.2 Hydration Tests 52
4.3.3 Leachate Compatibility 55
4.3.4 In-Plane Hydraulic Conductivity of a GCL 60
4.3.5 Differential Settlement 63
4.3.6 Desiccation Tests 69
4.3.7 Freeze Thaw 69
4.4 Effect of Freeze-Thaw in the Laboratory and Field 75
Chapter 5 — Manufacture and Deployment of GCLs 79
5.1 Perspective of CETCO with Respect to Bentomat ® 79
5.1.1 Overview 79
5.1.2 Summary of Technical Data 79
5.1.2.1 Low Normal Stresses 80
5.1.2.2 High Normal Stresses 80
5.1.2.3 Strain Softening ' 80
5.1.2.4 Long Term Effects ' 80
5.1.2.5 Testing Methods 80
5.1.2.6 Summary of Available Data 81
5.1.3 Modifications to the Clay Component 81
5.1.4 Future Needs for GCLs 81
5.2 Perspective of CETCO with Respect to Claymax® 82
VI
-------
5.2.1 Manufacture 82
5.2.2 Internal Shear Strength 82
5.2.3 Creep Shear Tests 82
5.2.4 New Products 83
5.3 Perspective of National Seal Co. with Respect to Bentofix® 83
5.3.1 Manufacture 83
5.3.2 Shear Tests 84
5.3.3 Creep Tests 84
5.4 Perspective of GSE Lining Technology with Respect to Gundseal® 85
5.4.1 GCL Product 85
5.4.2 GCL Field Test in Cincinnati, Ohio 86
5.4.3 Field Instrumentation of GCL/Bentonite Moisture 86
Monitoring Program
Chapter 6 — Report of ASTM Subcommittee D34.04 Subcommittee Activities 88
6.1 Physical Properties 88
6.2 Manufacturing QC/QA 88
6.3 Logistics 89
6.4 Endurance 89
6.5 Hydraulic Properties 89
6.6 Mechanical Properties 89
Chapter? — Regulatory Status of GCLs 90
7.1 Perspective of U.S. EPA Superfund Headquarters 90
7.2 Perspective of U.S. EPA Office of Solid Waste 90
Chapter 8 —- Panel Discussion 92
8.1 Critical Issues Concerning GCLs - EPA Project Team 92
8.1.1 Intimate Hydraulic Contact Vs. Shear Strength 92
8.1.2 Internal Shear Strength of GCLs under High Normal Stresses 92
8.1.3 Long Term Effects 92
8.2 Questions by the Audience 93
8.2.1 Do Bid Projects vs. Designs Need to be Sole-Sourced? 93
8.2.2 What is the Suitability of Replacing a Compacted Clay 93
Liner (CCL) in a Composite Liner with a GCL?
8.2.3 Should Designs for Waste Containment Structures Be 94
Based on Peak or Residual Shear Strengths?
Chapter 9 — References 95
vn
-------
Appendices
Appendix A •
Appendix B -
Appendix C -
Appendix D -
Appendix E -
- List of Attendees
- Total Down-Slope Displacement in Test Plots
- Differential Displacement in Test Plots
- Moisture Instrument Readings in Test Plots
- Results of Interface Shear Tests Performed at Drexel
University's Geosynthetic Research Institute
Appendix F — Evaluation of Various Aspects of GCL Performance, Prepared
by GeoSyntec Consultants
A-l
B-l
C-l
D-l
E-l
F-l
Appendix G — Summary of Bentomat Direct Shear Data, Prepared by CETCO G-1
Appendix H — Summary of Bentofix Shear Test Results, Prepared by H-1
National Seal Co.
vm
-------
List of Figures
Figure
2.1
2.2
2.3
2.4
2.5
•3.1
3.2
3.3
3.4
3.5
3.6
3.7
3.8
3.9
3.10
3.11
3.12
3.13
3.14
3.15
3.16
3.17
3.18
3.19
3.20
3.21
3.22
3.23
3.24
3.25
3.26
3.27
3.28
3.29
3.30
3.31
3.32
4.1
TJtifi.
General Configuration of Two Principal Types of GCLs.
Commercial GCLs Produced in North America.
Mohr-Coulomb Failure Envelope.
Curved Mohr-Coulomb Failure Envelope.
Peak and Residual Shear Strength.
Page No.
3
4
6
7
8
Layout of Test Plots. 12
Cross Section of Test Plots Containing a Geomembrane 13
(Plot Composition I).
Cross Section of Test Plots Not Containing a Geomembrane 14
(Plot Composition It).
Cross Section of Plot E. 14
Cross Section of Plots A, F, and P. 15
Cross Section of Plot I with Bentofix® I. 15
Cross Section of Plots D arid N with Bentofix® II. 16
Cross Section of Plots B and G with Bentomat. 16
General Cross Section of Plots on a 2H: 1V Slope. 17
General Cross Section of Plots on a 3H: IV Slope. 18
General Cross Section of Plots Showing the Width of a Plot. 18
Schematic Diagram of 2H: IV Test Plots 19
Schematic Diagram of 3H: IV Test Plots. 19
Anchor Trench Detail (Not to Scale). 22
Detail of Drainage at Toe for Sections with Geonet Drainage Layer 22
(Not to Scale).
Detail of Drainage at Toe for Sections with Granular Drainage Material 23
(Not to Scale).
Schematic Diagrams of Moisture Sensors. 24
Locations of Moisture Sensors and Extensiometers. 25
Location of Moisture Sensors in All Plots Except A and F. 26
Location of Moisture Sensors in Plots A and F. 26
Soil Types at 2H: IV Test Plots. 27
Calibration of Moisture Sensors for Soil A (Typical Calibration). 28
Calibration of Fiberglass Moisture Sensors with Bentonite. 29
Locations of Deformation Sensors. 29
Measurement of Shearing Displacement. 30
Attachment of Deformation Sensor to GCL. 31
Deformation Table at Crest of Slope. 31
Cross-Section at Crest of Slope Showing Cutting of Geosynthetics 33
Down to Mid-Plane of GCL on Test Plots with a Geomembrane.
Cross-Section at Crest of Slope Showing Cutting of Geosynthetics 33
Down to Mid-Plane of GCL on Test Plots without a Geomembrane.
Cutting of Slope with Gundseal®, Bentonite Side Facing Upward. 34
Cutting of Slope with Gundseal®, Bentonite Side Facing Downward. 34
Measured Water Contents in the Bentonite in Plot F. 38
Typical Plot of Shear Stress Vs. Displacement at a Normal Stress 51
of 17 kPa.
IX
-------
4.2
4.3
4.4
4.5
4.6
4.7
4.8
4.9
4.10
4.11
4.12
4.13
4.14
Results of Internal Shear Strength Tests on a GCL Hydrated for 51
Various Periods.
Rate of Wetting of Bentonite Component of Geomembrane- 53
Supported GCL Placed in Contact with Fine Sand at Different
Water Contents (from Daniel et al., 1993).
Rate of Wetting of Bentonite Component of Geomembrane- 53
Supported GCL Placed in Contact with Coarse Sand at Different
Water Contents (from Daniel et al., 1993).
Bentonite Sealed between Two HDPE Sheets in "Coupons." 54
Results of Hydration Tests on Bentonite Sealed in Coupons. 54
Plan View of Tank Used to Study Effect of Differential Settlement 64
on GCLs (from LaGatta, 1992).
Cross Sectional View of Tank Used to Study Effect of Differential 65
Settlement on GCLs (from LaGatta, 1992).
Definition of Distortion. 55
Tanks Used to Study Desiccation of GCLs (from Boardman, 1993). 70
Typical Results after Desiccated GCLs Were Permeated 71
(from Boardman, 1993).
Plan View of Cooling Coil Used for Freeze-Thaw Tests 72
(from Hewitt, 1994).
Cross Section of Tank Used for Freeze-Thaw Tests (from Hewitt, 1994). 73
Cross Section of Tank Used for Freeze-Thaw Tests (from Hewitt, 1994). 73
-------
List of Tables
Table
3.1
3.2
3.3
3.4
3.5
4.1
4.2
4.3
4.4
4.5
4.6
4.7
4.8
4.9
4.10
4.11
Title
Components of the GCL Field Test Plots.
Summary of Test Plots.
Moisture and Deformation Data.
Possible Pathways of Hydration of Plot F.
Geosynthetic Erosion Control Products.
Summary of Single-Point Direct Shear Tests.
Factor of Safety Vs. Hydraulic Head
Strength Parameters from Shear Testing.
Composition and Characteristics of the Permeant Liquids.
Summary of Results of Hydraulic Conductivity Tests with
Different Permeant Liquids (from Ruhl, 1994).
Comparison of Transmissivity Values Found in the Study by
Harpur, Wilson-Fahmy, and Koerner (1993) and the
In-Plane Hydraulic Conductivity Tests.
Results of Settlement Tests on Geosynthetic Clay Liners.
Summary of Experimental Results for 0,1, and 3 Freeze-thaw Cycles.
Results of Laboratory Freeze-Thaw Tests (n = Number of Freeze-
Thaw Cycles).
Results from the Field Freeze-Thaw Tests in Pans.
Hydraulic Conductivities of Large Specimens Removed from Lagoons.
Page No.
20
21
36
39
41
44
49
52
56
57-58
62
66-68
74
76
77
78
XI
-------
ACKNOWLEDGMENTS
This workshop was part of a larger effort on a cooperative agreement CR-815546-01-0
dealing with field performance of waste containment systems. The principal investigators for the
project are Robert M. Koerner (Geosynthetic Research Institute, Drexel University), David E.
Daniel (University of Texas), and Rudy Bonaparte (GeoSyntec Consultants). Several individuals
Srovided key support for the project, including George M. Koerner (Drexel University), John J.
owders (University of Texas), and Majdi Othman (GeoSyntec Consultants).
Robert E. Landreth, formerly with the EPA, as well as the EPA project officer, David A.
Carson, provided key support for this project and for and input to the workshop. The workshop
would not have been possible without the active participation of the many individuals who
participated in the workshop.
XH
-------
CHAPTER 1
INTRODUCTION
The first geosynthetic clay liner (GCL) workshop sponsored by the U.S. Environmental
Protection Agency (EPA) was held on June 7-8, 1990, in Cincinnati, Ohio, to discuss the use of
geosynthetic clay liners (GCLs) in landfill liner and cover systems. One of the primary
objectives of the workshop was to identify the most pressing research needs concerning GCLs.
The 75 attendees of the workshop identified two key research needs: (1) development of field
performance data, and (2) development of better information concerning the shear strength of
GCLs (Daniel and Estornell, 1991).
A second GCL workshop was held on June 9-10, 1992, at the same location. The
purpose of the second workshop was to present and to discuss new information developed on
GCLs. A variety of issues were discussed by the 175 attendees, but the key technical concerns
were once again the development of more field performance data and development of more
information concerning the shear strength of GCLs (Daniel and Boardman, 1993).
In the period since the first two GCL workshops, much has been learned about GCLs and
their use in waste containment facilities. A third GCL workshop was held in Cincinnati, Ohio, at
EPA's National Risk Management Research Laboratory on August 9 and 10, 1995. The third
Workshop was attended by 170 people (see Appendix A for list of attendees). This report
summarizes the proceedings from the workshop.
The 1995 GCL workshop had several objectives. The primary objective was to describe
the progress made on a research project in which GCLs were installed on test field plots at a site
in Cincinnati, Ohio. The purpose of constructing and monitoring the test plots was to evaluate
the mid-plane (internal) shear strength of GCLs in real-world, field-scale situations. A second
objective of the workshop was collect and disseminate the latest information that has been
developed on GCLs. Significantly more information is currently available on the performance
of GCLs compared to the time period of the previous workshops.
The primary presentations at this workshop constitute the major chapters of this report
and were as follows:
1. Background on GCLs (presented by David Daniel).
2. History and performance of the GCL field test in Cincinnati, Ohio (presented by David
Daniel, John Bowders, and Heather Scranton).
3. Recent research on GCLs (presented by Robert Koerner, Rudy Bonaparte, John Bowders,
David Daniel, Jason Kraus, David Carson, and Heather Scranton).
-------
4. Developments by manufacturers (presented by Robert Trauger [Bentomat®], John Fuller
[Claymax®], John Siebken [Bentofix®], and Richard Erickson [Gundseal®]).
5. Current Status of the development of ASTM standards for GCL testing (presented by
Larry Well).
6. Regulatory Status of GCLs (presented by Kenneth Skahn and Allen Geswein).
7. Panel Discussion of current issues and further research regarding GCLs and comments
from conference participants (presented by Rudy Bonaparte, David Daniel, and Robert
Koerner).
-------
CHAPTER 2
BACKGROUND ON GCLs
2.1 Types of GCLs
A geosynthetic clay liner (GCL) consists of bentonite sandwiched between two
geotextiles or attached to a geomembrane with an adhesive. Figure 2.1 shows cross-sections of
the two general types of GCLs. Presently, there are four manufacturers of GCLs in North
America and two in Europe. The ownership of geosynthetic manufacturers can change, as can
the product lines; therefore, the manufacturer should be consulted concerning specific products.
Cross-sections of the current GCLs are shown in Figure 2.2. The GCLs used in the field test
plots in Cincinnati include Gundseal®, Bentomat®, Claymax® 500SP, and Bentofix®.
(A) Geotextile-Encased GCL
(Bentonite Sandwiched between Two Geotextiles)
(B) Bentonite Supported GCL
(Bentonite Glued to Geomembrane)
Figure 2.1 General Configuration of Two Principal Types of GCLs.
Bentomat® and Bentofix® consist of dry sodium bentonite sandwiched between two
geotextiles. One geotextile is nonwoven while the other geotextile can be either woven or
nonwoven. The entire assembly is needle punched together.
-------
Bentofix®and Bentomat®
Woven or Non-Woven
Geotextile
Needlepunched
Fibers X.
Non-Woven Geotextile
Claymax®200R
Woven Geotextile
Woven Geotextile
Clay max® 500SP
Woven Geotextile
ftSlNftftftW
Woven Geotextile
Sewn Stitches
Gundseal®
Sodium Bentonite Mixed
with an Adhesive
Geomembrane
Figure 2.2 Commercial GCLs Produced in North America.
-------
Claymax® employs two woven geotextiles to contain the bentonite, which is mixed with
an adhesive. The GCL can either be unstitched (200R) or stitched (500SP). In the stitched
material, the rows of stitches are spaced 100 mm apart.
Gundseal® consists of sodium bentonite mixed with an adhesive and bonded to a
supporting geomembrane. The geomembrane is normally a 0.3- to 0.5-mm thick high density
polyethylene (HDPE) geomembrane, but virtually any geomembrane (smooth or textured) can be
used.
2.2 Advantages of GCLs
There are many attractive reasons for using GCLs as a component in a liner system or
cover system. First, GCLs typically have a hydraulic conductivity of 1 to 5 x 10"9 cm/s or less,
which makes them very effective as hydraulic barriers. In addition, the installed cost of GCLs,
which is about $5/m2, or $0.50/ft2, is low compared to compacted clay liners, particularly if clay
must be shipped from off site or if bentonite must be blended with soil to form the clay liner
material. Under ideal circumstances, with clay readily available on site and little or no
processing of the clay required, the cost of a 600-mm thick compacted clay liner can be as little
as $2/m2 to $3/m2 ($0.20/ft2 to $0.30/ft2) in 1995 dollars. If clay must be hauled great distances,
or if bentonite must be blended with on-site soils, the cost of a compacted clay liner is typically
$10/m2 to $40/m2 ($l/ft2 to $4/ft2). In addition to the material cost savings, the value of
airspace must often be factored into the cost analysis. For example, if the airspace that would be
occupied by a compacted clay liner becomes available for disposal of waste, additional cost
advantages are realized by the GCL. For instance, if the cost of waste disposal is $55/Mg ($50
per U.S. ton), and waste has a density of 890 kg/m3 (1500 lb/yd3), then the value of airspace
associated with a 600-mm thick compacted clay liner is $28/m2 ($2.80/ft2). Thus, the economic
incentives for using GCLs increase greatly when replacement of a compacted clay liner with a
GCL generates additional air space that is available for waste disposal.
There are other advantages of GCLs. GCLs are manufactured in rolls that can be shipped
anywhere. A local supply of clay is not needed if a GCL is used instead of compacted clay.
Installation of GCLs is easy and rapid. GCLs are resistant to freeze-thaw, wet-dry cycles, and
differential settlement. There are significant disadvantages of all liner materials, including
GCLs. The pros and cons of GCLs are described by Daniel and Boardman (1993)
These advantages are very significant and are the principal reasons why many owners
and designers of landfills and impoundments are increasingly selecting GCLs instead of more
conventional low-permeability compacted soils.
-------
2.3 Shear Strength of GCLs
One disadvantage of GCLs is the low shear strength of the bentonite contained within the
GCL. Shear test data on unreinforced, hydrated GCLs show friction angles of about 10 degrees.
In the two previous EPA workshops on GCLs, the shear strength of the bentonite in GCLs,
which controls the internal shear strength of unreinforced GCLs, was cited as a primary technical
concern in the use of GCLs in waste containment systems. The main factors affecting the
internal shear strength of GCLs include the magnitude of normal stress, water content of the
bentonite, rate of shearing, response of reinforcement, amount of deformation, and effects of
seismic loading. These factors are briefly reviewed in succeeding subsections,
2.3.1 Magnitude of Normal Stress
The classical Mohr-Coulomb failure criterion for the shear strength of soil is:
= c+atan
(2.1)
where t is the shear stress, c is the cohesion, a is the normal stress, and is the angle of internal
friction. The concept is illustrated in Fig. 2.3.
CO
<5
£
CO
Normal Stress (CT)
Figure 2.3 Mohr-Coulomb Failure Envelope.
The ideal Mohr-Coulomb failure envelope is linear. However, the relationship between
shear stress and normal stress for bentonite is not always linear (Fig. 2.4).
-------
1
Assumed Linear FaSure Envelope
Linear Failure Envebpe
Gives Correct Strength for
the Normal Stress of Interest
\
Actual, Curved
Failure Envelope
I
Normal Stress
of Interest
Normal Stress (a)
Figure 2.4 Curved Mohr-Coulomb Failure Envelope.
2.3.2 Water Content
The shear strength of bentonite is sensitive to water content. The angle of internal
friction decreases with increasing water content. For example, shear tests that were performed
on an unreinforced GCL at The University of Texas showed that at a water content of 20%, the
angle of internal friction was 22°, but when the water content was increased to 50%, the friction
angle of the unreinforced GCL decreased to 7° (Daniel et al., 1993). Hydrated bentonite is
significantly weaker than dry bentonite.
2.3.3 Rate of Loading
The rate of loading of GCLs affects the shear strength of the GCL. The general
experience with bentonite is the slower the loading, the lower the internal shear strength of the
GCL (Daniel et al., 1993). Thus, care should be taken in testing GCLs not to shear the GCL too
quickly. The procedure for determining shearing rate that is described in ASTM D3080 for
drained direct shear testing of soil is recommended.
2.3.4 Reinforcement
Many commercial GCLs are reinforced to enhance the internal shear strength of the GCL.
The reinforced GCLs used in the field test plots included Bentomat®, Claymax® 500SP, and
-------
Bentofix®. When a reinforced GCL is sheared internally, the needlepunched fibers or sewn
stitches are put into tension as shearing occurs, which enhances internal shear strength. However,
there are limitations on the benefits of this reinforcing, as discussed below.
2.3.5 Amount of Deformation
The peak shear strength is the maximum shear strength measured during shear.
Typically, however, many materials "strain soften" after the peak strength is reached. The
residual shear strength is the minimum post-peak shear stress, which typically occurs at a very
large displacement compared to the displacement at which the peak strength is generated. Figure
2.5 illustrates the difference between peak and residual shear strength.
If a reinforced GCL is loaded to very large shearing displacements (e.g., 50 mm or more),
reinforcing fibers may pullout from one or both of the geotextiles, break, or creep. If the
reinforcing fibers fail, the strength of the reinforced GCL may be about the same as that of an
unreinforced GCL. The key issue is how much deformation will actually occur in the field, and
whether there is a risk of residual conditions actually developing.
«
(/>
£ J
CO I
c3
CO
Stress-Strain Curvt?
A (Peak)
— • B (Residual)
Deformation
CO
CO
£
CO
O>
ra
CO
Mohr-Coulomb
Failure Envelopes
Normal Stress, a
Figure 2.5 Peak and Residual Shear Strength.
-------
2.3.6 Seismic Loading
Data on effects of cyclic loading on the internal shear strength of GCLs is very limited.
Several cyclic direct simple shear tests have been performed at The University of Texas. The
tests indicate that cyclic loading causes a slight increase in the shear strength of dry GCLs. The
increase in strength is the result of a slight densification of the dry bentonite during cyclic
loading. However, the tests indicate that saturated bentonite may undergo a slight reduction in
strength from cyclic loading. The reduction increases with increasing number of cycles of
loading. Results have not yet been described in a written report.
2.4 Interface Shear Strength
The interface shear strength of the GCL with an adjacent material can be the most critical
(i.e., lowest) shear strength. The shear strength of GCLs at interfaces can be affected by several
factors. One factor is the interfacing materials. For example, the friction angle between a GCL
and subsoil will be different from the friction angle between a GCL and a geomembrane. Also, a
textured geomembrane will typically have a higher interface friction angle with a GCL than a
smooth geomembrane. A second factor affecting the interface shear strength of geotextile-
encased GCLs is the different types of geotextiles used in making GCLs. An interface involving
a woven geotextile may have a lower shear strength than an interface involving a nonwoven
geotextile. A third factor is the degree of hydration of the bentonite and its potential mobility
through the geotextile components of the GCL. If hydrated bentonite can swell through the
geotextiles, the hydrated bentonite may "lubricate" the interface with an adjacent material. In
addition, the level of normal stress and amount of deformation can influence interface shear
strengths. Distortion before or during deformation could also be a factor. Finally, the amount of
deformation can influence interface shear strength. The large-displacement interface shear
strength is generally less than the peak value.
-------
CHAPTER 3
FIELD TEST PLOTS IN CINCINNATI, OHIO
Cooperative Agreement CR-821448-01-0 is a multi-task, multi-year research project
entitled, "Field Performance of Waste Containment Systems." The project is being conducted by
the U.S. EPA, Drexel University's Geosynthetic Research Institute, The University of Texas, and
GeoSyntec Consultants. One of the tasks on this project is to develop more information
concerning the shear strength of GCLs. The field test plots that are described in this section of
the workshop proceedings constitute the main effort on the GCL shear strength task of the
project.
The main objective in constructing the field test plots was to investigate the internal (mid-
plane) shear strength of GCLs in carefully controlled, field-scale tests. Other objectives were to
verify that GCLs in landfill cover systems will remain stable on 3H: IV slopes with a factor of
safety of at least 1.5, to monitor the deformation and creep of GCLs in the field for as long as
possible, to develop information on erosion control materials, and to better understand the field
performance of GCLs as a component in liner and cover systems.
Fourteen test plots have been constructed at the ELDA Landfill in Cincinnati, Ohio. Nine
of the plots were constructed on 2H:1V slopes and five were constructed on 3H:1 V slopes. Each
plot is about 9 m wide by 20 or 29 m long and is covered by approximately 0.9 m of cover soil.
Instrumentation was placed in each test plot (with a few exceptions) in order to monitor the
moisture content of the subsoil and deformations of the GCL. An additional plot consisting only
of cover soil was constructed on the 2H:1V slope. This plot did not contain geosynthetic
materials and was used as a control plot to study the effect of erosion on the cover soil on a plot
that did not contain any synthetic erosion control material.
Slope angles of 2H:1V and 3H:1V were selected to test the shear strength limits of the
GCLs. The rationale for selecting these slope inclinations was as follows. Many landfill final
covers have slopes of approximately 3H:1V. If GCLs are to be widely used in landfill covers,
they will have to be stable at a slope angle of 3H:1V. Thus, the 3H:1V slope was selected to be
representative of a typical landfill cover. However, it is not sufficient to demonstrate that GCLs
are stable on 3H:1V slopes — it must be shown that they are stable with an adequate factor of
safety. Many regulators and design engineers require that permanent slopes have a minimum
factor of safety for static loading of 1.5.
For an infinite slope in a cohesionless material, with no seepage, the factor of safety (F)
is:
10
-------
F =
tan<])
(3.1)
tan
where is the friction angle and (3 is the slope angle. If a GCL remains stable on the 2H:1V
slope, the friction angle of the GCL (assuming zero cohesion) must be at least 26.6°, and for this
friction angle, the factor of safety on a 3H:1V slope must be at least 1.5. Thus, the logic was to
try to demonstrate a minimum factor of safety of 1.5 on 3H:1V slopes, and in order to do this, it
was necessary to test the GCLs on 2H:1V slopes. It was recognized that constructing a 2H:1V
slope was pushing the test to (and possibly beyond) the limits of stability, not necessarily of the
mid-plane of the GCLs but certainly at various interfaces within the system.
3.1 Expectations Concerning Field Performance at the Beginning of the Project
During the conception and design of the field test plots, there were several expectations
concerning the performance of the GCLs. First, it was assumed that if the GCLs were placed
with the bentonite in contact with the subgrade soils that the bentonite would hydrate by
absorbing water from the adjacent soils. However, it was also assumed that if a geomembrane
separated the bentonite component of the GCL from the underlying subsoil, and a geomembrane
was placed over the bentonite to encase the bentonite between two geomembranes, that the
bentonite would be isolated from adjacent soils (except at edges) and would not hydrate.
A key expectation was that none of the GCLs would fail at the mid-plane on either the
3H:1V or 2H:1V slopes. This expectation was based on the results of mid-plane laboratory
shearing test on fully-hydrated GCLs. Interface shear slides were viewed as possible, but the
greatest concern was with the GCL/subsoil interface. It was predicted that deformations of the
GCLs would be downslope with the largest deformations on the 2H:1V slopes. Creep of the
GCLs was considered possible. Differential (shear) deformations were expected to be nominal.
3.2 Layout of the Test Plots
The test plots were constructed at the ELDA Landfill in Hamilton County, Cincinnati,
Ohio. Fourteen test plots containing a GCL as a component were constructed. The layout of the
plots is shown in Figure 3.1. Each plot was assigned a letter. Five plots (plots A-E) were
constructed on a 3H:1V slope, and nine plots (plots F to L, N, and P) were built on a 2H:1V
slope. An additional plot, plot M, which consisted of only cover soil and no geosynthetics, was
an erosion control plot that was installed, on a 2H:1V slope to document the degree of erosion
that would occur if no synthetic erosion control material was placed over the cover soil. In all
11
-------
other plots, a synthetic erosion control material covered the surface of the test plot. Plots on the
2H:1V slope were about 20 m long and 9 m wide; plots on the 3H:1V slope were about 29 m
long and 9 m wide.
3H: 1V Test Plots
DDDDD
ABODE
2H:1V Test Plots
'C?
H/W
N
Figure 3.1 Layout of Test Plots.
3.3 Plot Compositions
Four different types of GCLs were placed at the site: Gundseal®, Bentomat®,
Claymax® 500SP, and Bentofix®. Two styles of Bentofix® were employed. Bentofix® I
contained nonwoven geotextiles on both surfaces. Bentofix® II contained a woven geotextile on
the side that faced downward and a nonwoven geotextile on the side that faced upward.
Two general designs were employed. The principal design involved a subgrade overlain
by a GCL, textured geomembrane, geotextile/geonet/geotextile drainage layer, and 0.9 m of
cover soil. This cross section is typical of many final cover systems for landfills being designed
today. The geotextiles were heat-bonded to the geonet. A nonwoven, needlepunched geotextile
was used between the textured geomembrane and geonet in an effort to develop a high
coefficient of friction between the geomembrane and drainage layer. This type of plot
composition (Plot Composition I) is shown in Figure 3.2. Acronyms used are GT (geotextile),
GN (geonet), GM (geomembrane), and GCL (geosynthetic clay liner).
12
-------
Geosynthetic Erosion
Control Material
Cover Soil
GT/GN/GT
GM
GCL
Subsoil
rr.* t\* • .* • .• *\* • V • .• • .* • .* • .* • .*
•'•:-'::'s::''&:'':\'-'-:'''-^
Plot Slope GCL
A 3:1 Gundseal
B 3:1 Bentomat
C 3:1 Claymax
D 3:1 Bentofix II
Hoi Slope
F
G
H
I
N
P
2:1
2:1
2:1
2:1
2:1
2:1
GCL
Gundseal
Bentomat
Claymax
Bentofix I
Bentofix II
Gundseal
Notes:
1. Bentomat was installed with woven geotextile facing upward.
2. Bentofix I was Bentofix NW, which has a nonwoven geotextile on both sides.
3. Bentofix II was Bentofix NS, which had a woven geotextile facing downward.
Figure 3.2 Cross Section of Test Plots Containing a Geomembrane (Plot Composition I).
The second design involves a GCL overlain by 0.3 m of drainage soil, a geotextile, and
0.6 m of cover soil. This design is also typical of current GCL designs for final cover systems in
which a geomembrane is not used. This design (Plot Composition n) is shown in Figure 3.3.
In Plot E (Figure 3.4), the bentonite part of the Gundseal® GCL is in direct contact with
the subsoil, the geomembrane portion of the GCL is facing upward, and a geonet and 0.9 m of
cover soil has been placed on top of the GCL. Plots A, F, and P (Figure 3.5) were constructed
with the bentonite side of Gundseal® facing upward and the geomembrane portion in contact
with the subsoil. The composition of plots with Bentofix I and Bentofix II are shown in Figures
3.6 and 3.7, respectively, and plots with Bentomat are shown in Figure 3.8.
13
-------
Geosynthetic Erosion
Control Material
Cover Soil
Geotextile
Granular
GCL
Subsoil
Plot _Slope GCL
J 2:1 Bentomat (Woven Side Up)
K 2:1 Claymax
L 2:1 Bentofix I (NonWoven on both Sides)
Figure 3.3 Cross Section of Test Plots Not Containing a Geomembrane (Plot Composition II).
Cover Soil
GT/GN/GT _
(Drainage Layer)
GCL: Gundseal_
"Bentonite Down""
Subsoil
ILJiUiLJiLJiLILI£Jdi!J£Ji!JiLiattel
oncnnnnnnnnonnc
Figure 3.4 Cross Section of Plot E.
14
-------
Cover Soil
GT/GN/GT
(Drainage Layer)
GCL: Gundseal
"Bentonite Up"
Subsoil
Figure 3.5 Cross Section of Plots A, F, and P.
Cover Soil
GT/GN/GT
GM
nonwoven
GCL: Bentofix I -
nonwoven
Subsoil -
t-ILJLJLJlLJLJlLllLllLJlllLllLllLILJ
nnnDcnnnnonnncn
Figure 3.6 Cross Section of Plot I with Bentofix® I.
15
-------
Cover Soil
GT/GN/GT •
GM
nonwoven
GCL: Bentofix II
woven
Subsoil —
]nncnnnnnnnnnnnnnt
Figure 3.7 Cross Section of Plots D and N with Bentofix® II.
Cover Soil
GT/GN/GT
(Drainage Layer)
snnnnonnnnnnnncnnc
woven —*-t]
GCL: Bentomat
nonwoven
Subsoil
Figure 3.8 Cross Section of Plots B and G with Bentomat.
16
-------
Plot M is the erosion control section and consists only of 0.9 m of cover soil. There are
no geosynthetic materials or instrumentation inside the erosion control section.
General cross sections are shown in Figure 3.9 for plots constructed on the 2H:1V slope
and in Figure 3.10 for plots constructed on the 3H:1V slope. A general cross section of a plot's
width is shown in Figure 3.11. Each plot width was equal to two GCL panels minus a 150 mm
overlap. The spaces between plots on the 2H:1V slope ranged between 0 m and 1.5, and were
typically 1.5 m on the 3H:1V slope. There were graded drainage swales only on the 3H:1V
slopes. Table 3.1 lists the slope angles, plot, type of GCL, and a description of the plot cross-
section from top to bottom. Table 3.2 lists the composition, dimensions, etc., of each plot.
Figures 3.12 and 3.13 are schematics of the 2H:1V and 3H:1V slopes, respectively, at the site.
Subsoil
Drainage 1
(GT/GN/GT or Granular)
GM
Figure 3.9 General Cross Section of Plots on a 2H:1V Slope.
3.4 Anchor Trenches
Anchor trenches were constructed at the crest of each test plot. On the 3H:1V and 2H:1V
slopes all of the geosynthetic materials - GCL, geomembrane, and geonet (if present) - were
brought into the anchor trench. A membrane cap strip was placed over the GCL in the anchor
trench with the purpose of preventing moisture from entering the GCL from the crest of the plot.
A typical anchor trench detail is shown in Figure 3.14.
17
-------
Crest
Geocom posit e
Drainage
Material
0.9 m
Figure 3.10 General Cross Section of Plots on a 3H:1V Slope.
0.9m
Dra'nage Layer:
GT/GN/GT
or Granular
Erosion
Control
Mat
Coyer Soil
One GCL Panel
Width: 4 to 5.5 m
Two GCL Panel Widths: 8 to 11 m
Total Width of Test Plot: 10 to 13 m
Figure 3.11 General Cross Section of Plots Showing the Width of a Plot.
18
-------
No Geosynthetics
Granular Drainage
w/ GM & GT/GN/GT & No GM
Crest -*
Gundseal
-------
Table 3.1 Components of the GCL Field Test Plots.
Plot
A
B
C
D
E
F
G
H
I
J
K
L
M
N
P
GCL
Gundseal
Bentomat
Claymax
Bentofix n
Gundseal
Gundseal
Bentomat
Claymax
Bentofix I
Bentomat
Claymax
Bentofix I
Erosion
Control
Bentofix n
Gundseal
Target
Slope
(deg.)
18.4
18.4
18.4
18.4
18.4
26.6
26.6
26.6
26.6
26.6
26.6
26.6
26.6
26.6
26.6
Actual
Slope
(deg.)
16.9
17.8
17.6
17.5
17.7
23.6
23.5
24.7
24.8
24.8
25.5
24.9
23.5
22.9
24.7
Cross-section
(from top to bottom)
Soil/GN/GM/GCL (Bent, up)
SoiVGN/GM/GCL (W up)
SoiVGN/GM/GCL
Soil/GN/GM/GCL (NW up)
Soil/GN/GCL (Bent, down)
Soil/GN/GM/GCL (Bent, up)
SoiVGN/GM/GCL (W up)
SoiVGN/GM/GCL
Soil/GN/GM/GCL (NW-NW)
Soil/GT/Sand/GCL (W up)
Soil/GT/Sand/GCL
Soil/GT/Sand/GCL (NW-NW)
Soil
SoiVGN/GM/GCL (NW up)
Soil/GN/GM/GCL (Bent, up)
where:
Soil = cover soil
GN = geonet
GM = textured geomembrane
GT = geotextile
GCL = geosynthetic clay liner
Bent up = bentonite side of Gundseal facing upward (GM against subgrade)
Bent, down = bentonite side of Gundseal against subgrade
W up = woven geotextile of GCL up, nonwoven side of GCL against subgrade
NW up = nonwoven geotextile of GCL up, woven side of GCL against subgrade
NW-NW = both sides of GCL nonwoven
Bentofix I is Bentofix NW, with a nonwoven geotextile on both sides
Bentofix II is Bentofix NS, with a woven geotextile facing upward.
20
-------
Table 3.2 Summary of Test Plots.
Plot
A
B
C
D
E
F
G
H
I
J
K
L
M
N
P
GCL
Type
Gundseal
Bentomat
Claymax
Bentofix n
Gundseal
Gundseal
Bentomat
Claymax
Bentofix I
Bentomat
Claymax
Bentofix I
Erosion
Control
Bentofix E
Gundseal
Geo-
memb.
Cap
(Y/N)
Y
Y
Y
Y
Y
Y
Y
Y
Y
N
N
N
N
N
Y
Drainage
Type
GN
GN
GN
GN
GN
GN
GN
GN
GN
Sand
Sand
Sand
Sand
Sand
GN
Slope
3H:1V
3H:1V
3H:1V
3H:1V
3H:1V
2H:1V
2H:1V
2H:1V
2H:1V
2H:1V
2H:1V
2H:1V
2H:1V
2H:1V
2H:1V
Slope
Length
(m)
28.9
28.9
28.9
28.9
28.9
20.5
20.5
20.5
20.5
20.5
20.5
20.5
20.5
20.5
20.5
Crest
Elev.
(m)
179.2
179.2
179.2
179.2
179.2
157.9
157.9
157.9
157.9
157.9
157.9
157.9
157.9
157.9
157.9
Toe
Elev.
(m)
170.0
170.0
170.0
170.0
170.0
148.7
148.7
148.7
148.7
148.7
148.7
148.7
148.7
148.7
148.3
Test
Plot
Width
(m)
10.5
9.0
8.1
9.1
10.5
10.5
9.0
8.1
9.1
9.0
8.1
9.1
7.6
9.1
10.5
Notes:
1. Bentofix I is Bentofix NW, with a nonwoven geotextile on both sides
2. Bentofix II is Bentofix NS, with a woven geotextile facing upward.
3. Bentomat was installed with the woven geotextile facing upward.
3.5 Toe Detail
At the toe of the slope the geomembrane and geonet were extended beyond the GCL in
the plots on the 3H:1V and~2H:lV slopes. The geomembrane and geonet were extended
(daylighted) approximately 1.5 m past the end of the cover soil as shown in Figure 3.15. The toe
was designed to provide no buttressing effect for the cover soil.
21
-------
Geomembrane
Cap Strip
GT/GN/GT
GM
\Anchor
^Trench
^Backfill
Figure 3.14 Anchor Trench Detail (Not to Scale).
GT/GN/GT
GCL
1 to 2 m Daylighted
GT/GN/GT
Figure 3.15 Detail of Drainage at Toe for Sections with Geonet Drainage Layer (Not to Scale).
22
-------
The drainage detail at the toe of the 2H:1V slopes with granular drainage was developed
to prevent clogging (Figure 3.16). In order to maintain stability of the drainage material, a 4.2-m
wide piece of geonet was placed on top of the last 1.5 m of GCL at the toe of the slope. The
geonet extended approximately 1.5 m past the end of the GCL. Sand was placed on the geonet
above the GCL to ensure hydraulic continuity between the sand and geonet. The sand extended
on the geonet about 0.5 m past the end of the GCL. Cover soil was then placed on the sand and
geonet. At the toe of the slope, the geonet was excavated and exposed to permit drainage of the
encapsulated sand layer.
GCL
= 1.5 m Daylighted
GT/GN/GT
Figure 3.16 Detail of Drainage at Toe for Sections with Granular Drainage Material (Not to
Scale).
3.6 Instrumentation
The objectives of the instrumentation for the field test plots were to monitor the wetting
of the subsoil and the bentonite in the GCLs, and to monitor deformations of the GCLs.
Moisture sensors were installed to verify that the bentonite was hydrated, or in the case of plots
A, F, and P, to verify that the bentonite was dry. Extensometers were installed to document the
23
-------
mid-plane shear and creep of the GCLs in each plot. As there was a limited budget, the
instrumentation was selected based on simplicity, low cost, and redundancy.
3.6.1 Moi sture Sen sors
Moisture sensors were installed in each test plot in order to assess the moisture conditions
impacting the bentonite within the GCLs. Two types of sensors were used in the project: a
gypsum block sensor and a fiberglass mesh sensor (Figure 3.17). The gypsum block sensors
were placed in the subsoil beneath the GCLs; the fiberglass sensors were placed within the
bentonite of the GCLs. Both sensors operate on a resistance basis. The fiberglass sensors
contain a porous fiberglass mesh embedded in two wire screens . The resistance to flow of
electric current between the two screens is dependent on the moisture present in the fiberglass
mesh. The resistance is measured and converted to moisture content by comparison with a
calibration chart. The calibration is a function of soil type and the constituents of the soil
moisture. The gypsum block sensors have two concentric spirals of wire between which
resistance of gypsum is determined. The electrical resistance of the gypsum is a function of the
moisture content of the gypsum. The resistance is measured using a digital meter manufactured
specifically to measure resistance for these sensors.
Gypsum Block
t
Fiberglass Moisture Sensor
40 mm
40 mm
I
25 mm
25 mm
Figure 3.17 Schematic Diagrams of Moisture Sensors.
The sensors were placed on the centerline of one of the two GCL panels at three locations
top, middle, and bottom - of each plot as shown in Figure 3.18. The sensors were installed 5.2
24
-------
m, 10.7 m, and 16.8 m from the crest on the 2H:1V slope and 6.1 m, 15.2 m, and 24.4 m from
the crest on the 3H:1V slope. At each location two, and in some cases three, moisture sensors
were placed in the subsoil, at the subsoil-GCL interface, and in a few instances, above the GCL.
The purpose of the sensors was to monitor the moisture content of the bentonite and soil adjacent
to the bentonite. Because most plots contained a geomembrane above the GCL, placing sensors
in the cover soil would not provide information on moisture conditions within or near the GCLs.
Therefore, moisture sensors were generally placed adjacent to or beneath the GCLs. A cross
section of the moisture sensor installation in all plots except for plots A and F, is shown in Figure
3.19. Figure 3.20 shows how the moisture sensors were installed in plots A and F.
(MIDDLE)
• Cluster of Moisture Sensors
B Extensometer
111
2n
aS
4 IB
«
5IH
CREST
•1
• 2
B3
• 4
IDE "5
T \
Left GCL Panel Right GCL Panel
Figure 3.18 Locations of Moisture Sensors and Extensiometers.
The moisture sensors in Plot P were installed differently than the other plots. Only fiberglass
moisture sensors were installed in Plot P. Sixteen moisture sensors were placed in a 4 x 4 grid
on the upper side of the bentonite of the GCL but underneath the overlying geomembrane.
The gypsum blocks and digital meter were obtained from Soil Moisture Equipment
Corporation of Santa Barbara, CA. The fiberglass sensors were obtained from Techsas, Inc. of
Houston, TX.
25
-------
GT/GN/GT
GM
GCL
Gypsum
Block
Figure 3.19 Location of Moisture Sensors in All Plots Except A and F.
.•v.'V.-v.''V-.--v^^^
A^^•^^VA/^^^AVAV^V^V^V^^^^^^^^^^^^^VO/
Fiberglass
Gypsum
Block
Figure 3.20 Location of Moisture Sensors in Plots A and F.
26
-------
As mentioned above, the electrical resistance of a moisture sensor is measured and
converted to moisture content by comparison with a calibration chart. The moisture sensor
readout device used on this project reads from 0 to 100, with 0 corresponding to no soil moisture
and 100 corresponding to a very wet soil. However, the calibration is a function of soil type.
There are generally four different soil types at the site. The different subsoils on the
2H:1V slope are shown in Figure 3.21. Soil A is a gray fat clay, soil B is a clayey silt, and soil C
is a silty clay (field classifications). The subsoil on the 3H:1V slope is primarily a clayey silt
(soil D).
Crest
Soil A: Gray Fat Clay
Soil B: Clayey Silt
Soil C: Silty Clay
B
B
B
B
B
B
f
B
'C
B
C
'
C
C
>
C
A
•
A
Toe
PLOT: F G H I
K
M N
Figure 3.21 Soil Types at 2H: IV Test Plots.
Calibration tests were performed for both the gypsum block and fiberglass moisture
sensors for soils A, B, C, and D. A 1000 ml beaker was filled with soil, and a circular piece of
Gundseal® was placed above the soil with the bentonite portion of the GCL in contact with the
soil. A small layer of sand was placed over the GCL and a pressure of 18 kPa was applied to the
specimen. A gypsum block was inserted within the subsoil, and a fiberglass moisture sensor was
placed at the interface of the GCL and the subsoil. The subsoil was incrementally wetted, and
after the moisture gauge reading had equilibrated, the resistance reading was recorded and a
sample of the soil was obtained for measurement of water content. A typical calibration curve
for the gypsum block in the subsoil and the fiberglass moisture gauge at the soil/GCL interface is
shown for soil A in Fig. 3.22.
27
-------
j=p 100
o
V)
I
w
80
60
40
20
0
• Fiberglass Sensor-GCL Interface
-Gypsum Sensor Embedded in Soil
0.00 5.00 10.00 15.00 20.00
Water Content (%)
25.00 30.00
Figure 3.22 Calibration of Moisture Sensors for Soil A (Typical Calibration).
The calibration of the fiberglass moisture sensor with bentonite was performed as
follows. A fiberglass sensor was sandwiched between two prewetted pieces of Gundseal® so
that the sensor was surrounded by bentonite. Sand was placed below and above the GCLs, and a
pressure of 18 kPa was applied. After the moisture gauge reading had stabilized, the moisture
gauge reading was recorded. The calibration curve for the fiberglass moisture sensor with
bentonite is shown in Fig. 3.23. The scatter is due to the use of 15 different sensors in the
development of the calibration curve (each moisture sensor should ideally have its own
individual calibration curve). This calibration curve can clearly be used to qualitatively
distinguish whether the bentonite is relatively dry or saturated. Beyond that, however, statistical
scatter limits resolution. For example, a moisture gauge reading of 20 indicates that the water
content of the bentonite could range between 40 and 150%, and for a gauge reading of 80 the
water content of the bentonite could range between 190 and 290%. However, a gauge reading of
close to 0 clearly indicates that the bentonite is dry, and a reading close to 100 clearly indicates
that it is wet.
3.6.2 Deformation Gauges
Deformation gauges, or extensometers, were installed in each plot to measure
deformations and to assess shear strains in the GCL at multiple locations. Twenty deformation
gauges were installed in each plot (10 pairs on each panel). Five gauges in each panel were
attached to the upper side of the GCL along the centerline. Five gauges of each panel were
attached to the lower portion of the GCL directly opposite the gauge attached to the upper
surface of the GCL. The layout of the deformation gauges on the upper side of a GCL is shown
28
-------
in plan view in Fig. 3.24. With gauges on the upper and lower side of the GCL, the difference in
total deformation between the upper and lower gauges provides a measure of shearing
deformation. Figure 3.25 shows the measurement of the differential movement (AL) of the upper
and lower geotextile of the GCL.
O)
'o
o>
°^
o
V)
CO
2?
u
'o
120-
100-
80-
60-
40-
20-
0 •
0
TOO
150
200
250
300
Water Content (%)
Figure 3.23 Calibration of Fiberglass Moisture Sensors with Bentonite.
Crest
Toe
Figure 3.24 Locations of Deformation Sensors.
29
-------
GM
Bentonite
Deformatio
Cable
Fish Hook
Geotextile
8
Q.
CO
b
Upper Gage
Lower Gage
Time
Figure 3.25 Measurement of Shearing Displacement.
Each extensometer consisted of a wire running from its point of attachment to above the
crest of the slope. The wire was contained within a 6.4-mm OD (outside diameter) plastic
tubing, and was connected to a fish hook at the end of the wire (Figure 3.26). The fish hook was
attached by epoxy to the surface of the geotextile component of the GCL. Each wire extended
from the fish hook to a monitoring station, or deformation table, at the crest of the slope. A
deformation table is shown in Figure 3.27.
30
-------
i.4-mm OD Plastic Tubing
y3-mm Diameter Steel Cable
/ Epoxy,
Top Surface
Upper Geotextite
Bentonite
GCL
Figure 3.26 Attachment of Deformation Sensor to GCL.
Deformation
Indicator
Figure 3.27 Deformation Table at Crest of Slope.
31
-------
3.7 Construction
Construction of the plots began on November 15, 1994, and was completed on November
23,1994. The construction sequence was as follows:
1. Subgrade preparation.
2. Installation of moisture sensors in the subgrade and at the surface of the subgrade.
3. Placement of GCL.
4. Installation of the extensometers and deformation cables.
5. Installation of moisture gauges within the GCL (plots A, F, P).
6. Placement of geomembrane (not applicable to plots J, K, L, and M).
7. Placement of geonet or granular drainage layer (plots J, K, L, M).
8. Placement of geotextile (plots J, K, L, M only).
9. Placement of cover soil.
10. Construction of deformation tables.
3.8 Cutting of the Geosynthetics
With other geosynthetic material besides the GCL leading into the anchor trench, part of
the down-slope component of force created by the cover soil is carried by tension in the
geosynthetic materials. To concentrate all of the shear stress within the mid-plane of the GCL,
the geosynthetic materials above the mid-plane of the GCL were severed. The geosynthetics
above the mid-plane of the GCLs in plots A through D (3H:1V slope) were cut on April 13,
1995, and the geosynthetics above the mid-plane of the GCLs on the 2H:1V slopes and plot E
(3H:1V slope) were cut on May 2,1995.
In plots with geotextile-encased GCLs, the geonet, geomembrane, and the upper
geotextile of the GCL were cut at the crest of the slope down to the mid-plane of the GCL as
shown in Fig. 3.28. The geosynthetic materials in plots constructed with a granular drainage
layer were cut down to the mid-plane of the GCL as shown in Fig. 3.29. The granular drainage
material did not extend into the anchor trench, so the geotextile separating the cover soil and
drainage soil was cut as well as the upper geotextile in the GCL.
The cutting of anchor trench materials in plots with Gundseal® is shown in Figures 3.30
and 3.31. In the case with the bentonite side of the GCL facing up (Figure 3.30), the geonet and
geomembrane were cut leaving the entire GCL intact. In the case with the bentonite side of the
GCL facing downward (Figure 3.31), the geonet and the geomembrane of the GCL were cut.
32
-------
Upper Geotextile
in GCL Cut
GT/GN/GT
Figure 3.28 Cross-Section at Crest of Slope Showing Cutting of Geosynthetics Down to Mid-
Plane of GCL on Test Plots with a Geomembrane.
Upper Geotextile
in GCL Cut
GCL-
'.'. 'Subsoil •
Granular
Drainage
Figure 3.29 Cross-Section at Crest of Slope Showing Cutting of Geosynthetics Down to Mid-
Plane of GCL on Test Plots without a Geomembrane.
3.9 Field Performance
Deformation and moisture data have been collected once every two to three weeks since
the plots were installed. The data have been combined into three different types of plots. Two
different graphs are associated with deformation movements: 1) total displacement versus time,
33
-------
and 2) relative deformation versus time. The third type of graph is associated with the wetting of
the GCL and subgrade soils and shows the moisture gauge readings versus time.
Only GT/GN/GT
&GMCut
GT/GN/GT
GM
Gundseal GCL
(Bentonite Up)
Figure 3.30 Cutting of Slope with Gundseal®, Bentonite Side Facing Upward.
GT/GN/GT & GM
Component of GCL
Cut
GT/GN/GT
Gundseal GCL
(Bentonite Down)
Figure 3.31 Cutting of Slope with Gundseal®, Bentonite Side Facing Downward.
3.9.1 Deformation Data
3.9.1.1 Total Displacement Data
Total displacement graphs for all plots are presented in Appendix B. Total displacement
is the displacement of the gauge after the construction period. The total displacement is referred
34
-------
to as "steady state" displacement in Appendix B. Each total displacement vs. time figure
involves four plots. The two upper graphs show total displacement for the upper and lower
deformation gauges at the crest of the slope (gauge 1); while the two lower are for the upper and
lower deformation gauges for the toe (gauge 5) of the slope. Each graph includes data points for
both the left and right GCL panels of the plot. The vertical axis on the graphs show values of
increasing displacement down the axis (to simulate downslope movement). The accuracy of the
deformation gauges is thought to be approximately ± 25 mm. With this accuracy, the
deformation gauges can only be used to distinguish larger deformations or general trends in the
movement of the GCL. For plot A, the total, deformations associated with plot A are less than 25
mm.
3.9.1.2 Relative Displacement
The relative displacement, or the displacement of the upper gauge minus the
displacement of the lower gauge, has been plotted versus time in Appendix C for each test plot.
A positive value of relative displacement indicates that the GCL is experiencing shear in the
downslope direction. The vertical axis on the graphs in Appendix C shows values of increasing
displacement down the axis (to simulate downslope shear). The relative displacement versus
time graphs include pre-cutting as well as post-cutting deformations.
3.9.1.3 Displacements after Materials in the Anchor Trench were Cut
In both,the graphs of total displacement versus time (Appendix B) and relative
displacement versus time (Appendix C), the times when the anchor trench materials were cut is
noted. There was no sudden movement after the cutting of the geosynthetics in any test plot.
3.9.1.4 Moisture Gauge Readings
The readings from the moisture sensors are plotted versus time in Appendix D. These
figures provide a general indication of how the soil is wetting and to what extent the soil has
become saturated. The moisture reading is the instrument reading, not the actual water content.
3.9.1.5 Summary of Moisture and Deformation Data for all Plots
Moisture and deformation data for all test plots are summarized in Table 3.3. Only the
readings from the fiberglass moisture sensors are included in Table 3.3.
Plots A through E on the 3H:1V slope have all deformed less than 25 mm. The moisture
reading of the fiberglass gauges sandwiched between the bentonite of the GCL and the overlying
35
-------
geomembrane has remained zero for plot A (Gundseal® with the bentonite side facing upward)
and has ranged from 0 to 90 for most of the other plots on the 3H: 1V slope.
Table 3.3 Moisture and Deformation Data.
Plot
A
B
C
D
E
F
G
H
I
J
K
L
N
P
Slope
3H:1V
3H:1V
3H:1V
3H:1V
3H:1V
2H:1V
2H:1V
2H:1V
2H:1V
2H:1V
2H:1V
2H:1V
2H:1V
2H:1V
GCL®*
Gundseal (Bent, up)
Bentomat
Claymax
Bentofix II
Gundseal (Bent, down)
Gundseal (Bent, up)
Bentomat
Claymax
Bentofix I
Bentomat
Claymax
Bentofix I
Bentofix II
Gundseal (Bent, up)
Fiberglass
Moisture
Sensor
Reading**
0(B)
0-10
0-90
0-70
0-70
0 - 100 (B)
0-90
0-60
0-90
0-90
0-90
0-100
0-90
0-10(B)
Total
Movement
(mm)
<25
<25
<25
<25
<25
<375
sliding***
sliding***
<100
<50
<25
<50
<25
NA
where:
*Bent. up = bentonite side of Gundseal® facing upward (GM against subgrade)
Bent, down = bentonite side of Gundseal® against subgrade
**readout value (not water content); sensor placed at the interface of the GCL and subsoil
B = sensor placed within the GCL
*** interface slide occurred — see section 3.9.2
The plots on the 2H:1V slope have deformed more than the plots on the 3H:1V slope.
Plot F (Gundseal® with the bentonite side facing upward) has experienced deformations close to
375 mm. Plots G and H have slid at the geomembrane/GCL interfaces, and a discussion of the
interface slides will follow in the next subsection. The largest displacement of the other plots is
less than 100 mm. Deformations in plot P are not being monitored (no deformation gauges were
36
-------
installed). The moisture gauge readings show that all of the plots (with the exception of plot P)
have experienced wetting of the bentonite within the GCL, with a general trend of increased
wetting near the toe and crest of the slope. The wetting of plot F (Gundseal® with the bentonite
side facing upward) is discussed in section 3.10.
3.9.2 Interface Slides
There have been two slides at the interface between the GCLs and the overlying
geomembrane. Interface sliding occurred in plot H (Claymax® 500SP) on December 10, 1994
(21 days after the cover soil was placed), and plot G (Bentomat®) on January 12, 1995 (52 days
after the cover soil was placed). Both interface slides were associated with slippage at the
geomembrane-GCL interface (between a textured geomembrane and a woven geotextile), and in
both cases no mid-plane distress was noted. In both plots G and H, all of the material except the
GCL pulled out of the anchor trench (both slides occurred before the geosynthetics above the
mid-plane of the GCL were cut) and slid as one block down the slope leaving the GCL intact on
the slope. Both slides were sudden and rapid.
Analysis of the deformation data collected before the slides provided no indication that
sliding was imminent; the deformation readings were very small (< 25 mm of movement) prior
to sliding. The relative displacements, both construction and post-construction, in plot G do not
indicate that a significant amount of shear had developed within the GCL before the interface
slides occurred. Only construction deformations were collected for plot H prior to sliding.
Analysis of the readings of moisture sensors suggests that the crest and toe of plot G were
undergoing some hydration while the middle of the slope was still relatively dry. Moisture
readings from plot H show slight drying at the crest and middle of the slope, but the small
number of data points make drawing any conclusions difficult.
The objective of the test plots was to evaluate the internal, mid-plane shear strength of the
GCLs, not the interface strength of the various components. Although interface slides occurred
at these two test plots, the plots were viewed as successfully demonstrating the short-term,
internal, mid-plane stability of the GCLs, since there was no indication of internal distress or
deformation in the GCLs.
3.10 Plot F (Gundseal® with the Bentonite Side Facing Upward)
Plot F was constructed using Gundseal® with the bentonite sandwiched between two
geomembranes. Theoretically, the bentonite should remain dry because significant moisture
through the underlying and overlying geomembranes should be negligible and because wetting
along the edges and around any small holes in the geomembrane should be limited to a distance
37
-------
of a few millimeters or tens of millimeters from the source of the water. However, Plot F has
experienced wetting of the bentonite within the GCL.
When moisture sensors indicated wetting of the bentonite in the GCLs in early 1995,
holes were drilled through the cover soil, and core samples of the GCL were obtained to confirm
the moisture sensor readings and to investigate the cause(s) of the wetting. The results of the
water content tests are shown in Figure 3.31. The water content was found to be variable; some
areas were dry, and others were wet. The initial water content of the GCL in this test plot was
about 20%. Moisture contents of the samples ranged from 27 to 188%.
Top of Slope
Water .
Content x^
(%)
Initial Water
Content of
Bentonite « 25%
» '
27
s<"
42>
H ,
27
50 •
70.
*_^
— —.
34
31.
76 r
•
10
—^
32
|4'
M
86
•
i
I
104 |
\
F33
7
V
^^-^
28
28
188
65 - 1 25
57
, ^Deformation
Sensor
, Cables
Toe of Slope
r
Figure 3.32 Measured Water Contents in the Bentonite in Plot F.
Different pathways in which water could have hydrated the bentonite and the likelihood
of those pathways having occurred are listed in Table 3.4. There are three mechanisms that are
considered to be the most likely ones. One possible mechanism is through a surface leak at the
crest of the slope in which water traveled along the deformation cables that extended down the
entire slope length. Another possible pathway involves the "spider web" (a thin geotextile placed
over the bentonite in the GCL to keep excess bentonite from falling off the GCL) providing a
38
-------
wick for lateral water migration. A third possible pathway is lateral water movement from the
edges of the GCL through wrinkles in the geomembrane at the geomembrane/bentonite interface.
In addition to the unexpected hydration of the bentonite, Plot F has also undergone large
deformations in one of the GCL panels. The deformation gauges above the GCL have registered
large deformations (up to 375 mm) along the length of the slope in the right panel. The large
deformations could be due to disconnection of the gauges or to the wetting of the bentonite in the
GCL. Deformations in the left panel are comparatively small (generally < 1 mm). The highest
moisture contents were measured in the right panel (Figure 3.32).
Table 3.4 Possible Pathways of Hydration of Plot F.
Source of Water
Surface leak @ crest
Leak from edges along
slope face
Leak from below through
overlap seam
GCL was folded
longitudinally during
placement
Vapor transmission through
geomembrane
Installation damage - holes
& tears
Effect of the thin geotextile
used to secure bentonite
Wrinkles in geomembrane
provided seepage pathway
Comments
Path through slots cut in
upper GM for deformation
instrumentation cables
Water must travel « 2. 1 m
laterally to wet moisture
gauge and sampled area
Water must travel = 2. 1 m
laterally to wet moisture
gauge and sampled area
Folding GM could have
cracked or impaired the GM
Some vapor transmission is
expected
Cover soil on Plot F was
placed with a D-4 dozer.
No damage observed during
installation
Could GT provide lateral
flow path, increasing
transmissivity?
Any wrinkles between the
geomembrane and bentonite
would provide a pathway
for spreading of water
Likelihood that this Source
Caused Hydration of
Bentonite
Probable based on crest
conditions and surface
water drainage.
Tests on GCL without
geotextile indicate that this
mechanism is impossible;
but actual GCL has not
been tested
Tests on GCL without
geotextile indicate that this
mechanism is impossible;
but actual GCL has not
been tested
Unlikely
Rate of vapor transmission
is expected to be negligible
Unlikely
Possible that the geotextile
wicked water laterally
Some wrinkles are likely.
39
-------
3.11 Additional Test Plot P (Gundseal® with the Bentonite Side Facing Upward)
An additional test plot, plot P, was constructed (construction completed on June 13,
1995) in the same location that plot G (Bentomat®) had been located until the plot G slide. Plot
P was installed to verify that bentonite sandwiched between the geomembrane component of the
Gundseal® GCL and the overlying geomembrane could remain dry throughout the monitoring of
the test when placed on the 2H:1V. Plot P was installed similarly to plot F except that a spider
web was not used and deformation gauges were not installed. Sixteen moisture gauges were
installed in a 4x4 grid pattern to monitor the wetting of the bentonite. The moisture gauge
readings for Plot P have all been less than 10, indicating that the bentonite is relatively dry.
3.12 Erosion Control Materials
Erosion control materials were placed on the surfaces of all the test plots, except Plot M,
which was intentionally not covered with any erosion control material as a control plot. The
purpose of the erosion control materials was to stabilize the slopes rapidly and to maintain each
slope's surface integrity. Erosion control materials provide for the rapid growth of seeded grass
by retaining heat from the sun and limiting erosion due to overland runoff. The erosion control
materials give shelter to the seeds from flowing water and winds.
Table 3.5 summarizes the erosion control products that were placed on the various test
plots. Three plots (E, K, and N) had a sacrificial, biodegradable woody material applied to the
surface. All erosion control materials were installed according to manufacturer's specifications.
The erosion control materials were installed in an overlapping manner and stapled together and
to the soil at spacings of approximately 1 m. Some plots were seeded prior to placement of the
erosion control material, and others were seeded after the erosion control placement (depending
on the manufacturer's recommendation). The site owner provided for the seeding of the plots in
December, 1994.
Only visual information was available at the time of the workshop about the performance
of the various materials. It was noted that (1) all of the erosion control materials seemed to be
working well, and (2) there were significant erosion gullies in the control Plot M that did not
contain any erosion control material. As of August, 1995, the erosion control materials
appeared to be functioning as designed and have maintained the slopes' integrity.
3.13 Summary/Future Plans
The project so far has illustrated several important points about the use of GCLs as a
component in a waste containment system:
40
-------
1. GCLs have been stable on 2H: IV slopes with a factor of safety greater than or equal to
1.0 for mid-plane shear.
2. This indicates that GCLs on 3H:1V slopes have maintained a factor of safety greater
than or equal to 1.5 for mid-plane shear.
3. Reinforced GCLs have greater internal, mid-plane shear strength than the interface
strength between the woven geotextile component of the GCL and the textured
geomembrane that was used in constructing the test plots.
4. The weakest interface has been between a woven geotextile in the GCL and a textured
geomembrane.
5. In one plot the bentonite sandwiched between two geomembranes has become partly
hydrated, but the mechanisms of wetting remain uncertain.
Future work for this project will include continued monitoring of the instrumentation for
an additional 12 months, testing at the end of the project to determine water contents, etc., and
additional work on determining how the bentonite between two geomembranes can hydrate. A
final report will be issued in a little over a year.
Table 3.5 Geosynthetic Erosion Control Products.
Plot
A
B
C
D
E
F
G
H
I
J
K
L
M
N
P
Manufacturer
Tensar
Synthetic Industries
Synthetic Industries
Akzo
Akzo
Tensar
Tensar
Tensar
Synthetic Industries
Synthetic Industries
Akzo
Akzo
None
Akzo
Akzo
Product
TB 1000
Polyjute
Polyjute
Enkamat 7010
Enkamat 7010
TM 3000
TM 3000
TM 3000
Landlok 450
Landlok 450
Enkamat 7010
Enkamat 7010
Control Plot
Enkamat 70 10
Enkamat 7220
Color
Green
Beige
Beige
Black
Black
Black
Black
Black
Green
Green
Black
Black
-
Black
Black
Material
Polyolefin
Degradable Polypropylene
Degradable Polypropylene
Nylon
Nylon (with Excelsior)
Polyethylene
Polyethylene
Polyethylene
Polyolefin
Polyolefin
Nylon (with Excelsior)
Nylon
-
Nylon (with Excelsior)
Nylon
41
-------
CHAPTER 4
RESEARCH ON GCLs
4.1 Shear Behavior of GCL Interfaces at the EPA Test Plots in Cincinnati. Ohio
As discussed in Chapter 3, two test plots at the EPA GCL test facility in Cincinnati, Ohio,
have slid at the interface between the geotextile component of the GCL and the overlying
textured geomembrane. After these interface slides occurred, there was concern about potential
interface slides at similar interfaces in other test plots. To assist in evaluating if the interfaces
between materials at the test site would be stable, Drexel University's Geosynthetic Research
Institute (GRI) performed shear tests on critical interfaces.
At the time the shear testing was performed, thirteen plots had been constructed. Five
plots were constructed on a 3H:1V slope, and eight plots were constructed on a 2H:1V slope.
The shear testing focused on the plots and interfaces installed on the 2H:1V slope because these
were the more critical slopes. The interfaces of concern were:
1. The interface between the top of the GCL and the overlying textured geomembrane (T-
GM).
2. The interface between the top of the GCL and the overlying sand.
3. The interface between the bottom of the GCL and the underlying subgrade
(particularly the clayey subgrade soils identified as Soil A and Soil C in Fig. 3.21).
In addition, the internal shear strengths at the midplane of the GCLs were also evaluated through
direct shear testing, although no results were available at the time of the workshop.
The shear tests were performed according to ASTM D5321 in a 300 mm (12 in.) square
t
shear box. The specimens were hydrated for 2 to 10 days in the shear device under a normal
stress of 18 kPa. This stress is the approximate normal stress acting on the GCLs in the 2H:1V
test plots. The specimens were sheared at a strain rate of 1 mm/min. For each test, the peak and
large-displacement strengths were reported. Single point failure envelopes were created by
fitting a straight line through the origin and the failure point.
The interface between the top of the GCL and the textured geomembrane was the main
focus of the program of laboratory shear testing because of the two interface slides at the test
site. To simulate the field conditions as best as possible, site-specific products were used in the
tests. In order to obtain the large-displacement strengths, the specimens were sheared to
displacements of 35 mm.
The results of the direct shear tests are presented in Appendix E. The testing results are
summarized in Table 4.1. Table 4.1 also summarizes the calculated factor of safety of the plots,
based in Equation 3.1.
42
-------
The testing of the Bentofix® - HOPE (textured) geomembrane interface for plot N was
performed with the slit film woven geotextile at the interface. In reality, plot N was constructed
with the nonwoven geotextile of Bentofix® at the interface with the textured geomembrane
(facing upward). Therefore, the shear test on Bentofix® is not representative of the actual
conditions at the site. Another note of importance is that the shear test on the interface between
the bentonite portion of Gundseal® and the overlying textured geomembrane (plot F) was
performed with the bentonite at the "as-received" water content. As discussed in Section 3.10,
the bentonite component of the GCL has become hydrated in some locations. Thus, the direct
shear tests on dry bentonite are not representative of the actual conditions in Plot F.
The direct shear tests that simulated interfaces on plots G and H produced friction angles
lower than the actual angle of the slope. Therefore, if the interface shear tests had been
performed before construction, e.g., as would be the case for a typical landfill design project/the
interface slides on plots G and H would have been designed, and the slope would have been
designed differently (e.g., different materials, flatter slopes, or internal reinforcement in the
slope). The interface shear tests on all the other interfaces indicate that the peak interfacial
friction angles are greater than the slope angles, indicating stable interfaces, which correlates
well with field observations.
Plots J, K, and L were installed with sand as a drainage layer directly overlying the GCL.
Placed above the sand was a nonwoven geotextile overlain by cover soil that was protected with
an erosion control geosynthetic. A geomembrane was not installed. The concern in plots J, K,
and L was the shear strength of the interfaces between the geotextile component of the GCL and
the overlying sand. Direct shear tests were performed on this interface. The results are
presented in Appendix E and are summarized in Table 4.1. The sand-GCL interfaces appear to
be stable at the 2H:1V slopes, but just barely.
The stability of the subsoil at the 2H:1V test plots was of concern because of clayey
nature and high water content of the soil. Direct shear tests were performed by GeoSyntec
Consultants on the interfaces between the different GCL materials and the subsoil. These tests
showed that the interfaces should be stable, but just barely. During construction the goal was to
spread sand over the top of the prepared subgrade to provide additional resistance at the interface
of the subsoil and the GCL. However, due to scheduling problems and weather delays, no sand
was spread over the prepared subgrade. Instead, the GCLs were placed directly on the subgrade,
which, on the 2H:1V slopes, frequently consisted of wet, clayey soil.
43
-------
aS
to
CO
bO
CO
•s
&
CO
t*-i f~
0 £>«••£
•S **-.! e3 ^ '
i»dS|
t? ^^ tiUD
ts £>§
fe ca co
CO
8
4J K
*c
eE
"3 (P^OO
lct^£
M-aJ-sb'S
orf *ir* **H ^3 *F
i r^ *H ^c \^r
C
ca *X3 'EJ) bb
^|
^ y
ID
- O
O »
GCLType
00 Q
. 1
O
o
1
k
8
i(
Claymax®
_,
•M
fx^
ON
°
T— 4
(1)
JO
CS
CO
OO
s'
^
CN
^
en
1
i>
Bentofix®
TL-NW
-H
VO
cs
^4
flj
2
12
CO
m
cQ
en
i— H
en
2
3
CO
5
o
§>
o
^!
,_,
T-H
"
^4
li
2
12
CO
CN
ii
en
CS
en
N
CO
-a
=
H
J
is
^
ON
0
es
en
CJ
2
a
CO
o
en
cs
^
vj
o
cs
s
D
D
/D
Bentofix®
TL-NS 1
r
H
— I
d
o
*o
PQ
C/3
1
tn
s
§
a
rt
J P
CO
3
=3
=3
u
1
s
e
"S" O
O 1}
1 •?
!> fli (D
l|i
.ti ?-> ||
If "°|
coOE^
4«>
4>
I
8 o
fllfl
•|^°
2 P K
O .fc -S M
^8 -SJB
§ K rS'S
Ii ^1
N actually installed with
ar test on Gundseal® at '
Thermal Lock
= nonwoven geotextile on
nonwoven and woven ge
S £ ii £ "I
Iff ftflg
z
44
-------
4.2 Aspects of GCL Performance of Composite Liners Containing GCLs
GeoSyntec Consultants has studied several aspects of GCL performance as a component
in a liner system. The studies they have performed include:
1. The field hydraulic performance of composite liners containing GCLs
2. The hydration of GCLs adjacent to subgrade soil layers
3. The causes of failure of a landfill cover system containing a GCL
4. Direct shear strength of hydrated GCLs at high normal stress
5. Interface shear strength between unhydrated GCLs and textured geomembranes at high
normal stress. ;
A detailed report of the these studies is presented in Appendix F. Key aspects of the work are
summarized in the succeeding subsections.
4.2.1 Field Hydraulic Performance of Composite Liners Containing GCLs
The performance of a GCL as a liner in the field can be evaluated from several points of
view. For example, a GCL can be evaluated based on its ability to limit advective transport
(leakage), to restrict diffusive transport, or to attenuate the movement of chemicals. The focus of
this study was to evaluate how well a GCL performs as part of a composite top liner. The work
involved monitoring flow from leakage detection systems (LDS) below top liners at 26 waste
management units.
Flow through the top liner is not the only source of LDS flow. Other sources include
construction water, compression water, consolidation water, and infiltration water. Construction
water is the drainage, of water (mostly rainwater) that infiltrates the leakage detection layer
during construction but does not drain to the LDS sump until after the facility starts to operate.
Compression water is water that is expelled from the LDS layer as a result of compression under
the weight of the waste. Consolidation water is water expelled from any clay component of the
top liner as a result of consolidation under the weight of the waste. Infiltration water is ground-
water that infiltrates through the bottom liner under an inward hydraulic gradient.
The methodology for evaluating the LDS flow sources in this study was as follows:
1. Identify the potential sources of flow based on design, climatic and hydrogeologic
setting, and operating history.
2. Estimate the flow rates from each potential source.
3. Estimate the time frame for flow from each potential source.
45
-------
4. Evaluate the potential sources of flow by comparing measured flow rates to calculated
flow rates at specific times.
5. Compare leachate collection and removal system (LCRS) and LDS chemical
constituent data to further establish the likely source(s) of liquid.
Results from the study are summarized in Appendix F. The average LDS flow rate in the
26 waste management units attributable to top liner leakage ranged from 0 to 50
liters/hectare/day (Iphd), with most values being less than about 2 Iphd. The average flow rate in
the primary leachate collection and removal system was much higher: 5,350 Iphd.
It was convenient, in analyzing the data, to define an "apparent" hydraulic efficiency of
the top liner. The apparent hydraulic efficiency, AE, is defined as:
- n LDS Flow Rate , inn
= (1 - LCRS Flow Rate > x 10°
(4.1)
The range and mean of apparent efficiencies of the composite liner during the initial, active, and
post-closure periods were calculated for the cells with a sand LDS. During the initial period the
AEmean = 98.60% and AErange = 91.84 to 100.00%; during the active period the AEmean =
99.58% and AErange = 97.50 to 100.00%; during the post closure period the AEmean = 99.89%
and the AErange - 99.55 to 100.00%. The AEs in the initial period are lower than in the active
and post closure periods because the flow in the initial period is affected to a large extent by
construction water. During post closure, all of the flow is assumed to be from leakage through
the top liner, and the AE is highest in this period. The range and mean of apparent efficiencies
of the cells with top liners and geonet LDSs during the initial period of operation are: AEmean =
99.96% and AErange = 99.90 to 100.00%. Based on the data collected in this study, a composite
geomembrane-GCL liner provides excellent hydraulic containment.
The field performance of composite liners containing GCLs in 26 waste management
units has been evaluated. The study has found that LDS flow rates attributable to top liner
leakage range from 0 to 50 liter/hectare/day (Iphd), or 0 to 5 gallons per acre per day (gpad), with
most values being less than 2 Iphd. The "true" hydraulic efficiency of the top liner is greater
than or equal to 99.9%. Composite top liners containing GCLs, when appropriately used as part
of an overall liner system, can provide a very high degree of liquid containment capability.
46
-------
4.2.2. Hvdration of GCLs Adjacent to Suberade Soil Layers
GeoSyntec Consultants measured the rate of hydration of three GCLs used at the GCL
test site in Cincinnati, Ohio. The GCLs were placed in direct contact with the subgrade soils
from the test site and were expected to fully hydrate.
The GCLs included in the hydration tests were Claymax®, Bentomat®, and Bentofix®.
The soil used in the study was the subgrade soil from the western side of the 2H:1V slope at the
GCL field test site (Soil C in Figure 3.21). The soil was classified as a low plasticity clay (CL)
with a liquid limit equal to 41% and a plasticity index equal to 19%. A standard proctor
compaction test was performed; the optimum moisture content (OMC) was found to be equal to
20%, and the corresponding maximum dry unit weight was equal to 16.7 kN/m3 (106 pcf).
The soil was tamped into 75-mm-diameter molds to a relative compaction of 90% and at
water contents ranging from -4% OMC to 4-4% OMC. The range in water contents were selected
as the driest and wettest conditions that the GCLs might encounter at the GCL test site. The
GCL was placed against the soil, covered with a geomembrane, and loaded to 10 kPa with a
loading platen.
Plots for each GCL of the water content of the GCLs versus hydration time for subsoil
water contents equal to OMC - 4%, OMC, and OMC + 4% are shown in Appendix F. The
results are consistent for each GCL. The water contents of GCL specimens placed against
subgrades with a moisture content equal to OMC - 4% were all close to 40% after 25 days. GCL
specimens placed against subsoils with moisture contents equal to OMC had water contents
ranging from about 55% to 65% after 70 days, and GCL specimens placed against soils with
water contents equal to OMC + 4% had water contents ranging from 70% to 100% after 70 days.
The effect of overburden pressure was evaluated for one GCL (Bentofix®) by varying the
overburden pressure from 5 kPa to 400 kPa. The water contents varied randomly over a small
range (46% to 52%), and there was no trend. Therefore, overburden stress does not appear to
have an effect on the hydration of a GCL when placed in contact with subsoil.
Tests were performed varying the height of the soil column in contact with the GCL from
50 to 200 mm thickness. GCL specimens had slightly lower water contents (~ 40%)when in
contact with thinner (« 50 mm thickness) subsoils. Water contents of GCL specimens increased
to about 54% when in contact with thicker subsoils, but for subsoil specimens with 100 mm or
greater thickness, the water contents did not vary with the thickness of the soil. This indicates
that a subsoil specimen length greater than 100 mm will provide representative results.
Several conclusions can be drawn from the rate of hydration study:
47
-------
for reinforced GCLs. The major conclusions are that internal shear strengths of GCLs must be
measured at the appropriate normal stress, displacement rate, and total displacement to obtain
representative results, and that GCL shear strengths should be described in terms of the
corresponding range of normal stress for which the values were determined.
4.2.5 Shear Strength of GCL-Geomembrane Interfaces
For a project located in southern California, GeoSyntec Consultants performed 14
interface direct shear tests on the interface between an unhydrated GCL and a textured HDPE
geomembrane. The tests were performed in a 300 mm by 300 mm direct shear box following
procedures in general accordance with ASTM D5321. Three GCLs were tested using a range of
normal stress of 350 to 1,920 kPa. Two shearing rates were used (0.016 mm/s and 0.0007
mm/s), and it was found that the slower shearing rate yielded interface friction angles that were 1
to 2 degrees lower. It was also found that the normal stress had a significant effect on the
interface friction angles, which were 5 to 10 degrees lower at a normal stress of 1,920 kPa than at
350 kPa. It was also observed that there are relatively wide variances in the degree of texturing
of geomembranes, even from a given manufacturer. Further details are provided in Appendix F.
4.3 University of Texas Research
4.3.1 Direct Shear Tests of GCL Bentonite
The internal shear strength of the bentonite within the GCL determined from laboratory
tests are dependent on different preparation conditions and shearing methods. The University of
Texas is performing direct shear tests on the bentonite within samples of Claymax® 200R to help
develop appropriate direct shear testing procedures for GCLs. Two variables of shear testing of
GCLs that are being studied are the hydration time and the shear rate.
Shear tests have been performed on GCLs after they have hydrated for various lengths of
time - 24 hours, 48 hours and 72 hours - while holding all other variables constant (shear rate,
normal stress, etc.). Figure 4.1 is a typical plot of shear stress vs. displacement for three shear
tests performed after 24 hours of hydration under a normal stress of 17 kPa and at a shear rate of
1 mm/min. Similar tests were performed on specimens hydrated for 48 hours and 72 hours. A
graph of the shear stress vs. normal stress plots for the 24, 48, and 72 hour hydration series is
shown in Figure 4.2. Regression was performed for each hydration series (24,48, and 72 hours
of hydration) with the cohesion assumed to be equal to zero, and the friction angles are listed in
Table 4.3. The friction angles from the three different hydration periods were similar, indicating
that the period of hydration does not significantly affect the shear strength of the bentonite within
a GCL.
50
-------
15 h
ctf
Q_
CO
CO
CO
CB
o
CO
a.
•Test No. 1
•Test No. 2
•Test-No, 3
345
Displacement (mm)
Figure 4.1 Typical Plot of Shear Stress Vs. Displacement at a Normal Stress of 17 kPa.
CO
O.
oo
rt
CD
00
35 H
30
25
20
15
10
5
0
• 24 hr hydration
D 48 hr hydration
» 72 hr hydration
regression through origin
10
15 20 25
Normal Stress (kPa)
30
35
Figure 4.2 Results of Internal Shear Strength Tests on a GCL Hydrated for Various Periods.
51
-------
Table 4.3 Strength Parameters from Shear Testing.
Test Series
(Hydration Time)
24 hour hydration
48 hour hydration
72 hour hydration
Angle of
Internal
Friction
(deg)
29.5
30.0
30.5
Cohesion*
(kPa)
0
0
0
Average
Final
Water Content
(%)
166
182
171
Note: * assumed c = 0
Shear rate = 1 mm/min
64 mm diameter shear box
Tests will be performed on hydrated specimens of GCLs where the only variable is the
shear rate. All of the testing will be performed in 64 mm diameter shear boxes.
4.3.2 Hvdrarion Tests
At low water contents, bentonite has a very large suction potential. At a water content of
20%, a common water content for manufactured GCLs, bentonite can have a suction greater than
80 bars (Daniel et al., 1993). This means that bentonite will hydrate when placed in contact with
a material that is even slightly wet. To study how fast the bentonite in a GCL hydrates when in
contact with another material, hydration tests have been performed on several GCL specimens.
Three tests have been performed where the water content of the bentonite is monitored
over time. The first test involves placing Gundseal® in contact with moist fine sand and
monitoring the water content of Gundseal® over time. This test is repeated for different water
contents of the fine sand, and the results are shown in Figure 4.3. The second test is similar to
the first test except that coarse sand was used instead of fine sand. The results of the second test
are shown in Figure 4.4. The third test involves placing sealed "coupons" of Gundseal® in
contact with moist sand and monitoring the water content of the sealed coupons over a long
period of time. The "coupons" are 450-mm square sections of the GCL with a geomembrane
placed over the bentonite portion of the GCL (Figure 4.5). The edges are sealed by welding
along the edges. The variation of the water content of the bentonite within the sealed coupons is
shown in Figure 4.6.
52
-------
Water Content of Sand = 1%
0 10 20 30 40
Time (days)
Figure 4.3 Rate of Wetting of Bentonite Component of Geomembrane-Supported GCL Placed
in Contact with Fine Sand at Different Water Contents (from Daniel et al., 1993).
I
o
*E
CD
CO
o
O
CD
IS
200
150
100
Water Content
of Sand = 2.8%
10 20 30 . 40 50
Time, days
60
Figure 4.4 Rate of Wetting of Bentonite Component of Geomembrane-Supported GCL Placed
in Contact with Coarse Sand at Different Water Contents (from Daniel et al., 1993).
53
-------
450
TOD View: mm
Bentonite:
HOPE
Geo-
membrane
Cross Section:
Extrusion
Weld
•*^.^.^t^^:"*."W^. •*- -V^^^.^.^^
undseal
r*- •*••*••*••*• ^ -^^^^^. -». -v'^.^ir^-N.^v'^r^.^v'N.^.^j
450 mm
Figure 4.5 Bentonite Sealed between Two HDPE Sheets in "Coupons."
^.o
o^
*•-•'
"c
"c
o
O
v_
•+-«
OS
ZO
20
I
15
10
5
4
n
.^Bentonite in GCL
™ I^B ••
•• ^3
L n
y= -0.001X + 17.490
"n
-
^ Sand Subgrade
"«
-------
The water content of the bentonite in the sealed coupons changed by less than 5
percentage points. The results indicate that vapor transmission is not significant, and that a GCL
sandwiched between two geomernbranes should not hydrate, unless there is a source of water
such as water infiltration through a defect in the geomembrane.
4.3.3 Leachate Compatibility
Research has been conducted by Ruhl (1994) to study the effect of various permeant
liquids on the hydraulic conductivity of GCLs. Seven permeant liquids were used in tests:
simulated municipal solid waste (MSW) leachate, simulated hazardous waste (HW) leachate,
actual MSW leachate, fly ash leachate, a strong acid (HC1), a strong base (NaOH), and tap water.
Characteristics of the permeant liquids are listed in Table 4.4.
The GCLs were tested in, a flexible-wall permeameter. In the initial phase of the testing
program, some GCL specimens were hydrated with tap water prior to the introduction of the
permeant liquid ("prehydrated" GCLs), while other GCL specimens were permeated directly
with the chemical solution or leachate without any water hydration prior to permeation ("non-
prehydrated" GCLs). Under field conditions, a GCL could be hydrated by water prior to contact
with the chemical or leachate, e.g., by absorbing water from the adjacent soil. Alternatively, in
other situations the first wetting liquid could be the chemical or leachate. Results from the initial
tests that were performed for this study were consistent with findings reported by Daniel et al.
(1993), who found that the critical (highest) hydraulic conductivity occurs when the GCL is not
hydrated with water prior to permeation with chemicals. Therefore, in an effort to concentrate
the research on the critical condition, attention was focused on testing non-prehydrated samples.
However, when the dry GCLs were permeated directly with the chemical solutions or leachates,
the permeant liquid typically flowed rapidly through the GCL, and dozens of pore volumes
flowed through the test specimens in a few minutes or hours. The investigators were concerned
that this was insufficient time for full reaction and swelling to occur in the bentonite. In response
to this concern, a third condition of hydration was used in place of the ''non-prehydrated"
condition for the remainder of the testing program. For the third condition of hydration, GCLs
were given an opportunity to "hydrate" with the specific permeant liquid for a 48 hour period of
exposure prior to initiating the permeation process. The 48 hour "hydration" with permeant
liquid was accomplished by introducing the permeant liquid directly into the dry GCL test
specimen that was set up inside the permeameter, but closing the outflow valve from the
permeameter for 48 hours to prevent permeation and to allow, time for reactions to occur.
Table 4.5 summarizes the results of the tests. Further detail is provided by Ruhl (1994).
55
-------
Table 4.4 Composition and Characteristics of the Permeant Liquids.
Permeant Liquid
Reason Liquid Was Used
Composition/Characteristics
Simulated Municipal
Solid Waste (MSW)
Lcachate
Designed to represent an aggressive
"worst case" MSW leachate, with a high
concentration of calcium (which is a
chemical known to increase the hydraulic
conductivity of bentonite).
0.15 M acetic acid
0.15 M sodium acetate
0.007 M salicylic acid
1000 mg/L calcium
pH = 4.4
(Based on the formulation for synthetic
MSW leachate by Stanforth et al., 1979)
Simulated Hazardous
Waste (HW) Leachate
Designed to represent an aggressive HW
leachate with relatively high
concentrations of organics in an aqueous-
phase leachate.
4000 mg/L acetone
2000 mg/L benzoic acid
3000 mg/L phenol
1000 mg/L methyl chloride
100-200 mg/L cadmium
(Formulation based on data published by
Bramlett, 1987; McNabb et al., 1987; and
Sai and Anderson. 1991)
Simulated Flyash
Leachate
Designed to simulate leachate from Type
C coal combustion fly ash.
Constituents not measured.
pH=11.5-12
(Leachate developed following procedure
of Texas Water Commission. 1985)
Actual MSW leachate
Selected as a typical, real-world MSW
leachate.
Key constituent concentrations are:
112 mg/L calcium
520 mg/L chloride
368 mg/L sodium
100 mg/L potassium
50 mg/L magnesium
4,340 mg/L sulfate
2.5 mg/L phenol
116 mg/L acetone
9 mg/L benzene
11 mg/L toluene
41 mg/L ethylbenzene
130 mg/L xylene
87 mg/L chlorobenzene
1,800 mg/L dissolved solids
60 mg/L suspended solids
687 mg/L COD
254 mg/L BOD (5-day)
312 mg/L total organic carbon
pH = 7
(Obtained from a midwestern US
municipal landfill, which had been in
operation for several years)
Tap Water
Designed to comply with
recommendations of ASTM D5084 for
permeation with tap water.
Key constituent concentrations are:
18 mg/L calcium
54 mg/L chloride
15 mg/L magnesium
25 mg/L sodium
Hydrochloric Acid
Designed to represent a strong, acidic
leachate.
0.1 M solution of reagent grade HC1 in
distilled water; pH = 1
odium Hydroxide
(NaOH)
Designed to represent a strong, basic
leachate.
0.1 M solution of reagent grade NaOH in
distilled water; pH = 13
56
-------
3
%3
"eg a>
B 9
ES
W
.S w o
If
.S §
fiu
•s
o 4> fe
(S 3 E
"o
"3 -s "c?
in
ffiu
d
o
>l (
o
E.
§
•0
1
Condition of I
T3
•3
0"
3
-*-»
e
s
VO
•*
^
a
CN
VO
9
Bentofix
1
c
^.
!^
1
|
^.
^
§
Os
>T)
p:
0!
o
C
k
4
1
^
,
CN
CN
t—
PJ
oo
Tfl
ft
c
c
S
CO
CO
«
Cf
Contamina
0
8
,
^j-
^
9
u
CN
•c
B
O
b.
S
c£
1
1
^
o
^,
<— '
1
"
Bentofix
b
^5
OO
1
s
73
i
i
i
§
§§
_
OS
oo
PJ
CN
Bentomat
1
1
J
i
'
3
P-H
i-H
^.
f~
o
[X
g
CN
ei
a
I
<0
00
co
>o
o
i
— 1
»
•a
c
O
c
^
c/;
a
pi
C
Contamina
i
i
„
§
VO
m
0
W
co
1
Regular Gi
vo
1—1
Os
S
TJ-
10
jj
vo
"e?
IK
•o
C
«
CO
V
c/
B
Contamina
1
1
c
O
'"H
g
1
o
•*
ti
CN
1
C
O
t_
E
4
%
^
/— *
f—H
II
UH
5
CO
co
^.
,
CN
OO
S
J.
T3
4)
•a
c
3
V
«
c/
B
Contamina
1
1
)
VO
oo
S
1
>n
CN
0
•c
G
O
w
5
t
=^
g
VO
r-
CN
i
"er
CO
•o
C
2
c7
c<
CO
4'
PS
Contamina
i
^
z
p
CO
oo
o
1
CN
oo
u:
1
c
Regular Gi
t
*
S
II
i
p
Os
CN
CO
00
•*
1
ir
"e?
4)
CO
•a
O
c
CO
c/
{/
p:
Contamina
1
<
•«
g
i
c-
—
tf
eg
|
c
a
u
£
>
57
-------
I
•a
I
a
CT"
og.
.11
fT. C3
r-^ ^^
w
•1
ill is
."3 H
•8
! 3 5:
i s E
O *tli
"•* >•
= '3 "to' H^;
" S 1 "~
T3 O.
O
u_
O
ir
o\
:
O
i
ti
CO
i
II
's.
a I
I
•5|3
>-J I ea
s Ip
S*
E
1
CO
58
-------
The significant findings from the work may be summarized as follows:
1. The hydraulic conductivity to tap water was 3 x 10'10 to 2 x 10'9 cm/s, which is typical
of other measurements that have been reported on GCLs.
2. The GCLs were permeated with a simulated MSW leachate that was rich in calcium
(1,000 mg/L Ca+2). When dry GCLs were permeated directly with the simulated MSW
leachate, the hydraulic conductivity was typically 10'6 to 10'5 cm/s, which is about 4
orders magnitude higher than the hydraulic conductivity to water. Replacement of
sodium by calcium is the obvious cause of the high hydraulic conductivity. The GCLs
were much less permeable to the simulated MSW leachate when they were hydrated
with water prior to permeation with leachate: the prehydrated GCLs had about the same
hydraulic conductivity to leachate as to water.
3. Very different results were obtained when the GCLs were permeated with a real MSW
leachate, which did not tend to increase the hydraulic conductivity of the GCLs. The
real MSW leachate appeared to be less aggressive than the simulated MSW leachate
because the real MSW leachate: (1) contained roughly equal amounts of monovalent
and polyvalent cations; (2) contained suspended solids and microorganisms that tended
to plug the pores of the bentonite; and (3) produced gases of decomposition that tended
to block flow paths.
4. Permeation of dry GCLs with a strong acid or base resulted in high hydraulic
conductivity (10~6 to Ifr7 cm/s). In nearly all cases, prehydration with water, followed
by permeation with the acid or base, resulted in hydraulic conductivities that were
several orders of magnitude lower than the same GCLs that were directly permeated
with the acid or base. Non-water-hydrated bentonite had no significant capacity to
buffer the pH of the acid or base, but for pre-hydrated GCLs, about 15 pore volumes of
flow were required before the pH of the effluent liquid became similar to that of the
influent liquid.
5. When the GCLs were permeated with simulated hazardous waste leachate, the
hydraulic conductivity was in the range of 10"9 to 10'10 cm/s (similar to and in some
cases less than the values for tap water). The simulated hazardous waste leachate
contained 10,000 mg/L of several organics, but this concentration was not high enough
to significantly alter the dielectric constant of the simulated leachate and, hence, there
was little impact on hydraulic conductivity.
6. Tests on simulated fly ash leachate showed little difference between the hydraulic
conductivity to water and the hydraulic conductivity to the fly ash leachate.
59
-------
7. In addition to these specific findings, several practical conclusions were reached that
should be considered by engineers or scientists who are evaluating the potential for a
chemical solution or waste liquid to increase the hydraulic conductivity of a GCL:
A. Small concentrations or organic liquids do not appear to be damaging;
B. Perhaps the most aggressive liquid is one that is free of suspended solids or
microorganisms, but which contains a high concentration of multivalent cations
such as Ca+2;
C. Simulated and real leachates may react very differently with GCLs — whenever
possible, real leachate should be used;
D. The first wetting liquid (water or leachate) is a critical testing variable — the
worst case is always direct exposure of a dry GCL to the chemical solution or
waste liquid;
E. The period of time that the GCL is exposed to the chemical solution or waste
liquid prior to permeation can affect the test results significantly — laboratory
tests should simulate field conditions as closely as possible;
F. Laboratory tests, particularly on GCLs that have first been hydrated in water,
may have to continue to a large number of pore volumes of flow (e.g., 15 for
strong acids or bases) in order to achieve full breakthrough of key chemical
constituents; and
G. Not all contaminant-resistant bentonites are as resistant to attack as regular
bentonites for certain chemicals — great care should be exercised when
selecting a contaminant-resistant bentonite to make sure that it is resistant to the
specific chemical solution or leachate that is of interest.
4.3.4 In-Plane Hydraulic Conductivity of a GCL
Testing has been performed to study the in-plane hydraulic conductivity and
transmissivity of GCLs in contact with geomembranes (Amble, 1994). To study the in-plane
hydraulic conductivity of a GCL, the GCL and geomembrane specimens were placed between
two acrylic half cylinders. The geomembrane was bonded to the face of one acrylic half
cylinder. The GCL and acrylic half-cylinders were oriented vertically in a flexible wall
permeameter. Therefore permeation of the GCL took place through the plane of the bentonite.
In tests with GCLs having geotextiles on both sides, one or both geotextiles were smeared with a
thin layer of bentonite. This was to study the effects that each geotextile had on the in-plane
hydraulic conductivity of the GCL.
60
-------
Hydraulic conductivity tests were performed on Bentomat® Regular, Bentomat® HS (a
product with adhesive added to one of the geotextiles to bond the needlepunched fibers to the
geotextile), Bentofix® (the version with a woven geotextile on one side and a nonwoven
geotextile on the other side), Claymax® 200R, Claymax® 500SP, Gundseal® with contaminant-
resistant bentonite, regular Gundseal®, and Gundseal® with a very light-weight fabric backing
on the bentonite. In some cases bentonite was placed on one or both of the geotextiles of the
GCL. Some of the GCLs have different geotextiles on either side of the GCL. Side A of a GCL
denotes the woven geotextile while side B denotes the nonwoven geotextile of the GCL. For
Claymax® 500SP only, side A indicates the cream-colored woven geotextile while side B
indicates the gray woven geotextile.
Results are provided by Amble (1994). It was generally found that the presence of a
geotextile increased the in-plane transmissivity of a GCL, and that the trahsmissivity decreased
with increasing confining stress. Table 4.6 summarizes the results and compares the findings
with those published by Harpur, Wilson-Fahmy, and Koerner (1993).
In general, geotextile-encased GCLs with bentonite smeared on the woven geotextile
(side A) had a higher in-plane transmissivity than when bentonite was smeared on the nonwoven
geotextile (side B). In other words, the nonwoven geotextile (side B) is more transmissive than
the woven geotextile (side A). Exceptions include regular Bentomat® and Bentomat® HS,
where side A was more transmissive than side B. The cream-colored, woven geotextile (side A)
in Claymax® 500SP appeared to be slightly more transmissive than the gray, woven geotextile
(sideB).
The three different samples of Gundseal® were found to have in-plane transmissivities
that varied by less than one order of magnitude. The lowest hydraulic conductivity was found
for the sample of Gundseal® with a light-weight fabric, indicating that the fabric component of
this GCL does not significantly affect its in-plane transmissivity. Amble (1994) provides further
detail.
Samples of Gundseal® typically had lower in-plane transmissivities than the geotextile-
encased GCL samples. In some cases, the differences were almost insignificant. However, in
other cases, the differences were as large as four orders of magnitude, indicating that the
geotextiles formed flow paths.
61
-------
Table 4.6 Comparison of Transmissivity Values Found in the Study by Harpur, Wilson-
Fahmy, and Koeraer (1993) and the In-Plane Hydraulic Conductivity Tests.
Description of GCL
Transmissivity (m2/s) at a
compressive stress of 69 kPa
(from Harpur, Wilson-Fahmy,
andKoerner, 1993)
Transmissivity (m2/s) at a
compressive stress of 69 kPa
(from Amble, 1994)
Bentofix® with bentonite on
Side A (leaving the nonwoven
geotextile to provide most of the
in-plane transmissivity)
(GCL "E" from Harpur et al.)
8xlO-n
4xlO-9
Bentofix® with bentonite on
Side B (leaving the woven
geotextile to provide most of the
in-plane transmissivity)
(GCL "B" from Harpur et al.)
9xlO-12
2xlO-n
Bentomat® with bentonite on
Side B (leaving the woven
geotextile to provide most of the
in-plane transmissivity)
(GCL"D")
IxlO-10
3xlO-9
Claymax® 200R with bentonite
on Side A (leaving the woven
geotextile on the other side to
provide most of the in-plane
transmissivity)
(GCL"C")
6x10-12
IxlO-11
Gundseal®
(GCL "A")
3xlO-12
2xlO'12
62
-------
4.3.5 Differential Settlement
Research has been performed at the University of Texas to study the effect of differential
settlement on the hydraulic conductivity of GGLs. More information about differential
settlements of GCLs is provided by LaGatta (1992). In the research, large sections of GCL were
bolted to a frame that was fitted to the inside of tanks measuring 2.4 m in length, 1.2 m in width,
and 0.9 m in height (Figure 4.7). The sections of GCL were placed over a large water-filled
bladder (Figure 4.8). The GCLs were buried beneath 600 mm of gravel, and a head of water of
300 mm above the top of the GCL was maintained. The deformations were created by slowly
deflating the bladder. The GCLs were tested with and without an overlap.
Four GCLs were tested. Bentomat® was tested with the nonwoven geotextile facing
downward. Gundseal® was tested with the geomembrane portion of the GCL facing upward.
Two versions of Claymax were tested: the original version of Claymax® and Claymax® 500SP.
In the cases where overlaps were tested, the overlaps were created 225 mm wide.
The GCLs were subjected to differential settlement in either a dry or saturated state. The
water-filled bladder was deflated in four stages for GCLs that were initially hydrated with water,
and in one stage for GCLs that were deformed in a dry state. Differential settlement may be
characterized by ,the distortion, D/L, which is defined as the settlement, P, over a horizontal
distance, L (Figure 4.9). The average tensile strain caused by distortion can be computed by
integrating over the deflected shape to determine the arc length of the deformed section from
simple mechanics. If tensile strains are large enough, the GCL may crack and undergo an
increase in hydraulic conductivity.
Outflow was collected from a line leading out from the base of the tank. Hydraulic
conductivity was calculated using Darcy's law. The thicknesses of the GCLs were determined
from separate tests performed in the laboratory using the appropriate vertical compressive stress.
The hydraulic gradient ranged from 20 to 30. Typical test lengths were about 4 months each.
The results of the tests are summarized in Table 4.7. Most of the GCLs maintained a
hydraulic conductivity ^ 1 x 10'? crn/s (a common regulatory limit) while subjected to a tensile
strain of 1% to 10% or more, depending on the material and conditions of testing. Some GCLs
were able to bridge over the underlying subsidence because of the tensile strength of the
geosynthetics of the GCL. Overlapped and non-overlapped materials performed about the same
mostly due to the swelling and self-healing capability of the bentonite. Further information is
provided by LaGatta (1992) and Boardman (1993).
In general, GCLs can undergo large settlements without experiencing a significant
increase in hydraulic conductivity. It appears that GCLs can withstand more differential
settlement than compacted clay, but less than very flexible geomembranes.
63
-------
C/5
M
i-J
O
JS
C
n
0
O
c
0
g
H
I
1
'o
C/J
*3
*?
BS
Table 4.7
"^
1
•n
3
§
u
.a
i
s
4«
?'!
£w
_^
s s
CM C2
sa"2
QCO
§
R
0
*2
CO
Is
II
1
1
•5
S
3
.S
- CS
O cs cs en
o 6 6 66'
..-.OS VO i-H Tt
c:' i— i TJ- vo r-
o t-i cs en -^-
I
»— i
©
S ^_^
O 0
+2
O
ON ON ON »~H
O O O O
X X X X
T-H CS 1— 1 OO
ON ON ON ON
6066
X X X X
CS CS CS *-i
o o o
6666
X X X X
i— 1 OO trH OO
en vo o
O O T- < VO
VO OO t—
C-- t- "*
O t-i en
0666
® T- C ^ OO
O T— i cs en
00
CS
o\
O
X
en
ON
O
X
en
o
6
X
VO
„
>o
en
en
en
6
•o
r-
'-*
S.
en
^^ G\ O^ ON O^
66666
X X X X X
en cs T— i cs cs
00066
X X X X X
TJ- en n
T— 1
f*- en en en
en en oo r—
O »-i en >r>
06666
° °° en o? ^
O »— i cs en ^
$
cs
oo
6
X
^
6
X
CS
6
X
cs
•>*
en
en
en
6
t2
o
£
en
66
-------
css
&
a,
i
T3
3
•rt
0
CJ
e
in
oo vo
6 6
X X
t-< CS
OO >O
6 6
X X
1-1 CS
00
X X
CS ~H
oo
o 6
en
0 0
o 2
o ^
oo
oj
CS
0
X
en
0
X
CS
00
6
X
oo
s
en
en
en
0
m
-
Q
(S
en
O\ ON t— oo in
66 6 6 6
X X X X X
en vo en \o es
6 o 6 6 6
X X X X X
en t^ cs vo cs
C5
•~i ON r- oo r-»
66666
X X X X X
OO Tt i— i IO O\
en os i-; t~;
o 6 6 cs* vo
r*^ en ^~ en
vo en O t-~
o i-i cs en
00000
^•^ T-H en ^st* oo
O, r- 1 cs en "«t
a?
!H
»— H
®°
o
v^
rs ^^
11
43^
ON
6
X
r-
0
X
ON
O
6
X
»n
en
en
en
O
£
-
Q
!Zi
CS
ON ON V)
6 66
XXX
ON Os. »— <
Os Os ir,
O O O
XXX
ON ON 'vf
ON ON ON
0 O 0
XXX
vo oo ON
r-H 0
o o cs"
en o
w-i O
O CS
000
£=5 CS CS
° T-H CS
O I-H CS
§5
IH
^H
_
e ^ cu
§ fB
.S 1(5
Ci ^^^ ^"""^
ON oo r^
O O O
XXX
f- i— i vo
op oo vo
666
XXX
I-H i-^ en
ON ON ON
0 O O
XXX
'd- CS 00
*-* vo
o 6 ^
IS
066
0 !U H
^^ T-I CS
O T-H CS
s
£*
cs
00
6
X
cs
oo
6
X
en
oo
6
X
cs
in
en
en
en
0
£
-
o
K
en
67
-------
en
I
J
o
o
O
a
o
•S3
oo
1
1
o
a
o
I
I
if
£co
•o §
e a
— I 1— 1
?^ r^ ?S rS ?C
en >o eS cs es
66666
W ^ ^ S*^ «s>
^ rS rS rS rS
*^J* ON cM co co
ON o\ t*1* oo r^
66666
X X X X X
CS ^ T-l r- 1 ON
es ON »-i »-<
o 6 6 cs en
T— 1
t^- en ^ oo
vo en o cs
O r- 1 CS >O
06666
O^ o »o 2
»— i en ^ ^
O »— ' cs en rf
CO
-
0.
@°^
^s 55
in QH
st
II
1
X
ON
oo
6
X
t— <
6
X
>n
>n
en
en
en
O
$
-
&
cs
1
^
1
'o
12
1
W-i
«
«n
en
VO
vo
cs
0
55
^
GO
-
•o 9
c 2
8
P.,
^
|
^
^
1
U-i
&
ON
es
en
oo
O
T— (
^
I/)
-
/— ^
ffi) O
ll
11
OS
o
6
I— I
X
en
0
en
6
X
rt-
en
en
en
0
£
*-
18
es
68
-------
4.3.6 Desiccation Tests
Research has been performed by Boardman (1993) to study the effect of cyclic wetting
and drying on the hydraulic conductivity of GCLs. Three GCLs were tested: Bentomat®,
Gundseal®, and Claymax®. Large sections of each type of GCL were bolted to a frame
(preventing contraction) that was fitted to the inside of tanks measuring 2.4 m in length, 1.2 m in
width, and 0.9 m in height. The overburden soil was 0.6-m thick. As indicated in Figure 4.10,
eight PVC pipes were installed vertically above the GCL; six of the pipes were used for the
injection of hot air, while two were used for extraction of air. A water head of 300 mm was
maintained throughout the permeation portion of the tests. The GCLs were not subjected to
multiple stresses (e.g., differential settlement and desiccation).
The GCL was first allowed to hydrate, then water was drained out of the tank. The
heating system was turned on and the GCL was dried for 2 to 3 weeks. The GCL became
cracked and desiccated. The GCL was then rehydrated by applying water to the surface of the
GCL at a rate of 40 mm per hour. When outflow through the GCL sample stopped after the
initial rehydration, the water head was raised in increments of 100 mm to 300 mm over several
days. The tests were carried out until steady-state flow was reached.
The variation of hydraulic conductivity with time for a typical GCL is shown in Figure
4.11. The geotextile-encased GCLs (Bentomat® and Claymax® 200R) swelled and self-sealed
upon re-hydration, after one cycle of wetting and drying. When the desiccated GCLs were re-
hydrated, water initially flowed rapidly through most of the desiccated samples, but the bentonite
quickly expanded and the hydraulic conductivity decreased as the cracked bentonite began to
adsorb water and swell. The long term, steady value of hydraulic conductivity was essentially
the same before and after the desiccation cycle. In tests performed on a GCL containing
bentonite attached to a geomembrane (Gundseal®), there was no outflow of water either before
or after the wetting and drying cycle. Due to the presence of the geomembrane, very little of the
GCL actually became hydrated, but the bentonite in the overlapped area did self seal. The initial
hydraulic conductivity after rehydration was high. If the GCLs are slowly wetted, unlike the
tests where 40 mm of water was added to the surface of the GCL per hour, the GCL would have
time to adsorb water and to swell without allowing seepage through the GCL.
4.3.7 Freeze Thaw
Research has been performed by Hewitt (1994) to study the effect of freeze-thaw on the
hydraulic conductivity of GCLs. The tests were performed using the same tanks described
earlier. The GCLs were covered with 25 mm of gravel, and a cooling manifold was placed on
the gravel. Figures 4.12,4.13, and 4.14 show the testing equipment and set-up.
69
-------
Plan View
Key:
O Hot-Air Injection Well
Vacuum Extraction Well
Piezometer
1 m
Air In
Cross-Sectional View
Air Out
Blower
if Heating Element
Figure 4.10 Tanks Used to Study Desiccation of GCLs (from Boardman, 1993).
70
-------
-5
CO
>N
J^
'•S
I
o
o
.J2
«
10
10-6
-7 •
ID
'9
10
Days After Initial Rehydration
Figure 4.11 Typical Results after Desiccated GCLs Were Permeated (from Boardman, 1993).
The GCLs were allowed to hydrate with a water level in the tank of 300 mm. When the
flow of water through the GCLs became steady, the water depth was lowered to 25 to 50 mm.
The GCLs were then frozen by circulating chilled fluid at -17° C through the cooling manifold.
Thermisters installed above and below the GCL in the gravel were used to monitor temperature.
The GCLs were chilled until the thermisters registered less than -1° C. This usually took 4 to 5
days. The fluid circulation was stopped, and the GCL was allowed to thaw under ambient
conditions. Thawing of the GCLs to 10° C usually took 3 to 4 days. Each GCL sample was
subjected to three cycles of freeze-thaw.
Flexible wall hydraulic conductivity tests were performed on each type of GCL to
determine the hydraulic conductivity. The hydraulic conductivity of the GCLs (in the large
tanks) after three freeze-thaw cycles can then be compared to the flexible wall hydraulic
conductivity. The tests performed in the large tanks are called bench-scale tests. Results from
the bench-scale and flexible wall tests are presented in Table 4.8.
71
-------
1.2m
E
•*
cJ
IP
\
9
3Sft£^
'Km^Simm^im
mymmmj^f^^.
Steel
Tank
Angle
Frame
Tee Connection
Hose
p^Bentonite
Seal
Pea Gravel
.Wood
"Frame
Wood
Support
Copper Tubing Serving as Manifold
for Circulation of Refrigerated Liquid
Figure 4.12 Plan View of Cooling Coil Used for Freeze-Thaw Tests (from Hewitt, 1994).
72
-------
Steel
Tank
Granular
Bentonite
Side Seal
Copper
Outlet Tube
03
o
Figure 4.13 Cross Section of Tank Used for Freeze-Thaw Tests (from Hewitt, 1994).
Return Flow for
Chilled Liquid
Copper
Tubing V
GCL
Temperature
Probes
Wooden
Blocks
Chilled Liquid
Flow Line
Ethylene
Glycql
Solution
Freezer
Cooling
Coils
Cooling
Coils
Figure 4.14 Cross Section of Tank Used for Freeze-Thaw Tests (from Hewitt, 1994).
73
-------
Table 4.8 Summary of Experimental Results for 0,1, and 3 Freeze-thaw Cycles.
GCL
Geotextile-
Encased,
Needle-Punched
GCL
Geotextile-
Encased, Stitch-
Bonded GCL
Geomenibrane-
Supported GCL
Type of Test
Bench-scale (Intact)
Bench-scale (Overlap)
Flexible-wall
Bench-scale (Intact)
Bench-scale (Overlap)
Flexible-wall (No Stitching)
Flexible-wall (With Stitching)
Bench-scale (Overlap)
Flexible-wall (Bentonite Only)
a
(kPa)
7.6
7.6
12.4
7.6
7.6
10.3
10.3
7.6
12.4
Hydraulic Conductivity (cm/s)
0 Freeze-Thaw
Cycles
2xlO-9
4xlO-9
2x10-9
SxlO-9
IxlO"8
2x10-9
2xlO-9
No flow(*)
3x10-9
1 Freeze-Thaw
Cycle
3 x lO-9
7 x 1C'9
3 x lO-9
5 x lO-9
1 x lO'5
4 x 10-10
2xlO-9
-
2 x lO-9
3 Freeze-Thaw
Cycles
2x10-9
3 x lO-9
3 x 10-9
7 x lO-9
7 x 10'6
3 x lO-9
2x10^9
No Flow(**)
2 x lO-9
Note: a = effective stress (kPa)
Not available
(*) No flow occurred during 25 days of permeation
(**) No flow occurred during 52 days of permeation
It is concluded that under the conditions of these tests, most GCLs (intact and overlapped
panels) can withstand at least 3 freeze-thaw cycles without undergoing a significant increase in
hydraulic conductivity. However, the tests were performed under carefully controlled conditions
in laboratory devices at a compressive stress of about 8 kPa. The GCLs were only subjected to 3
freeze-thaw cycles, and the conditions of freeze-thaw were not superimposed with other
environmental stresses (e.g., differential settlement and desiccation). Ultimately, field data are
needed to understand how GCLs perform in the field.
74
-------
4.4 Effect of Free.7.e-Thaw in the Laboratory and Field
The University of Wisconsin studied the effect of freeze-thaw on GCLs at laboratory and
field scale. Three GCLs were studied in the laboratory portion of the research: Claymax®
200R, Bentomat® CS, and Bentofix®. Three GCLs were used in the field portion of the
research: Claymax® 500SP, Bentomat® CS, and Gundseal® (results of Gundseal® are
discussed in Erickson et al., 1994).
In the laboratory tests, GCL specimens with a diameter of 150 mm were placed in
flexible-wall permeameters for saturation and hydraulic conductivity measurement (GRI test
method GCL-2 with an average effective stress of 14 kPa and hydraulic gradient of 75). After
permeation the specimens were removed from the permeameters and sealed in plastic freezer
bags to prevent desiccation and placed in a freezer. After freezing, the specimens were removed
and allowed to thaw at room temperature. If the specimen was to be permeated after thawing,
the specimen was placed in the permeameier and then allowed to thaw. The hydraulic
conductivity of each specimen was measured after 1, 3, 5, and 20 freeze-thaw cycles.
The field scale tests were performed first in lagoons. The lagoons were 8 m wide by 10
m long by 0.6 m deep. One lagoon was built for each type of GCL in 1992. The GCL was
placed in the lagoon and overlain by 0.25 m of gravel. A sump was used to measure outflow,
and thermisters were used to determine when the GCLs froze. The GCLs in the lagoons were
exposed to two freeze-thaw cycles (two winters of exposure with one freeze-thaw cycle per
winter). The field tests in lagoons developed problems with the seepage collection and
measurement system, and the tests within the lagoons were abandoned.
Smaller specimens (1 m x 1 m) were cut from the GCLs in the abandoned lagoon field
tests. Even smaller specimens were cut from the 1 m x 1 m sections and placed in flexible wall
permeameters for evaluation of hydraulic conductivity. In these tests, sidewall leakage tended to
be a problem and resulted in high measured flow rates.
Another field test was then constructed in 1993. Each type of GCL was placed in three
separate pans, 1 larger pan (1.34 m x 1.34 m) and 2 smaller pans (0.84 m x 0.84 m). The large
pans were constructed on welded HDPE, and the small pans were constructed on molded HDPE.
Thermisters were placed below the GCLs and were used to determine when freezing had taken
place. Gravel was placed over the GCL specimens. The large pans and one of the smaller pans
contained GCLs seamed per manufacturers' specifications. The second smaller pan in each
group contained a seamless GCL specimen. The GCLs in the pan field tests were allowed to
hydrate under 30 mm of water for one week. The water level was then raised to 250 mm, and the
GCLs were allowed to hydrate for one month. Tests were performed with the water level at 250
75
-------
mm. The hydraulic gradient ranged from 5 to 15 and averaged 10. Measurements of hydraulic
conductivity were made by measuring outflow, and the hydraulic conductivity was measured
before winter up to December 1993 and after winter starting April, 1993.
The laboratory freeze-thaw tests were performed on GCL specimens that were permeated
to a steady-state hydraulic conductivity. The results are listed in Table 4.9. All initial hydraulic
conductivities are less than 4.9 x 10'9 cm/s and all of the final hydraulic conductivities are less
than 3.3 x 10'9 m/s. The k2o/ko ratio is equal to the ratio of the hydraulic conductivity after
exposure to twenty freeze-thaw cycles to the initial hydraulic conductivity. All but one specimen
had a k20/ko ratio-less than one, and the one value that is greater than one is only slightly greater
(1.1). A t-test confirmed that the decrease in hydraulic conductivity is statistically significant.
Table 4.9 Results of Laboratory Freeze-Thaw Tests (n = Number of Freeze-Thaw Cycles).
Sample Number
Bentofix®-!
Bentofix®-2
Bentofix®-3
Bentomat®-!
Bentomat®-2
Bentomat®-3
Claymax®-!
Claymax®-2
Claymax®-3
Claymax®-4
Initial
Hydraulic.
Conductivity
koxlO-9
(cm/s)
2.9
4.9
5.6
3.1
3.1
2.9
3.8
2.9
4.2
4.9
Hydraulic Conductivity after n Freeze-Thaw
Cycles, kn x 10~9
(cm/s)
ki
3.0
1.6
1.7
2.9
1.7
1.8
2.9
2.4
3.5
4.1
k3
2.8
2.3
3.5
2.8
2.4
1.4
4.8
2.7
3.4
3.2
k5
NP
2.7
3.6
1.3
2.5
1.5
4.2
3.6
3.2
4.4
k20
3.2
2.2
2.5
1.7
1.9
1.9
3.4
2.1
2.4
3.3
k,n/kn
1.10
0.45
0.45
0.55
0.61
0.66
0.89
0.72
0.57
0.67
note: NP = not performed
The field freeze-thaw tests were performed in nine test pans with three pans per GCL
type. The three pans for each type of GCL included one large pan and two smaller pans. The
results of the field tests are listed in Table 4.10. The results for Gundseal® are reported by
Erickson et al. (1994). There was essentially no increase in hydraulic conductivity for four of the
76
-------
six Bentomat® and Claymax® specimens. One specimen had no seepage before winter so no
comparison could be made between the before- and after-winter hydraulic conductivities. The
large, overlapped Claymax specimen exhibited an increase in hydraulic conductivity by a factor
of 25 after exposure to one winter (one freeze-thaw cycle).
Table 4.10 Results from the Field Freeze-Thaw Tests in Pans.
Specimen
Bentomat®, 1.8 m2
Bentomat®, 0.7 m2
Bentomat®, 0.7 m2
Claymax®, 1.8 m2
Claymax®, 0.7 m2
Claymax®, 0.7 m2
Seam?
Yes
Yes
No
Yes
Yes
No
Before- Winter
Hydraulic
Conductivity
(cm/s)
1.5 x 10-8
1.0 x lO-8
no outflow
2.8 x lO-8
2.0 x 10-8
2.4 x lO-8
After-Winter
Hydraulic
Conductivity
(cm/s)
1.9 x 10-8
1.4 x lO-8
l.OxlO-8
7.0 x lO-8
3.0x10-8
2.8 x lO-8
Ratio of Value After
Freeze-thaw to
Value Before Freeze
Thaw
1.3
1.4
.
25.0
1.5
1.2
Laboratory hydraulic conductivity tests were performed on specimens taken from the
field, where the GCLs had been subjected to two freeze-thaw cycles. The results from the
hydraulic conductivity tests are summarized in Table 4.11. Low hydraulic conductivities were
measured for two 0.45 m diameter Claymax® specimens. Sidewall leakage occurred in two 0.45
diameter Bentomat® specimens. The Bentomat specimens were retrimmed to a diameter of 0.30
m and retested. One Bentomat specimen had a low hydraulic conductivity. The other continued
to exhibit sidewall leakage and was not tested.
77
-------
Table 4.11 Hydraulic Conductivities of Large Specimens Removed from Lagoons.
Specimen
Bentomat®
Bentomat®
Claymax®
Claymax®
Diameter
(m)
0.30
0.30
0.45
0.45
Hydraulic Conductivity
(cm/s)
not measured*
1.7x10-8
3.5x10-8
6.3x10-8
The laboratory and field freeze-thaw tests produced similar results. Freeze-thaw had
essentially no effect on the hydraulic conductivity of the GCLs. The only exception was the
large, overlapped Claymax® GCL tested in the field in a test pan. The hydraulic conductivity of
the overlapped Claymax® specimen was significantly more permeable after winter. The other
overlapped Claymax® specimen did not exhibit an increase in hydraulic conductivity. The
hydraulic conductivities of laboratory specimens exposed to freeze-thaw were approximately one
order of magnitude lower than the hydraulic conductivities of the field specimens exposed to
winter weather. This may be a result of testing conditions (higher effective stresses and
hydraulic gradients used in the laboratory).
The structure of the frozen and thawed GCL specimens was examined. Vertical and
horizontal sections were cut through the frozen GCL specimens. Small, randomly oriented
lenses of segregated ice were observed in both sections. The thawed specimens revealed no
cracking commonly seen in compacted clays exposed to freeze-thaw. The structure of the
thawed specimens appeared similar to the hydrated specimens that had not been exposed to
freeze-thaw. The specimens removed from the lagoons were also devoid of cracks induced by
freeze-thaw.
78
-------
CHAPTERS
MANUFACTURE AND DEPLOYMENT OF GCLs
SI Perspective of CETCO with Respect to Bentomat ®
Mr. Robert Trauger of Colloid Environmental Technology Co. (CETCO) presented an
overview of past and present needs in the GCL industry, a summary of the technical data
collected for Bentomat®, and a list of the future needs for GCLs in general.
5.1.1 Overview
Mr. Trauger identified the principal concerns for GCLs in 1992 as:
1. Intimate contact
2. Freeze/thaw
3. Seam performance
4. Shear strength
5. Puncture resistance
6. Conformance testing
7. Subtitle D equivalency
Many of the concerns from 1992 have been resolved or researched to a significantly
further degree than in 1992. According to Mr. Trauger, the concerns for GCLs in 1995 are shear
strength, chemical compatibility, and manufacturing quality control (QC). Shear strength has
emerged as the major technical concern, and the issues associated with shear strength are:
1. Testing techniques
2. High normal stress
3. Interface strength
4. Long term behavior (creep)
5. Seismic effects
5.1.2 Summary of Technical Data
Shear tests have performed on Bentomat® to examine some of the issues concerning the
shear strength of GCLs. The results of some of the shear testing were presented and a general
description of the conclusions from each test is provided. A summary of Bentomat® direct shear
results and the data from the direct shear and creep shear tests are located in Appendix G.
79
-------
compare test results and products. More workshops on GCLs would be helpful to everyone
involved, because everyone benefits from shared information.
5.2 Perspective of CETCO with Respect to Clavmax®
Mr. John Fuller of CETCO discussed the manufacturing process, internal shear behavior,
creep shear testing, and interface friction testing of the Claymax® 500SP (Shearpro®) product.
He also described some new products, namely the high tensile strength GCL.
5.2.1 Manufacture
During manufacture of Shearpro®, bentonite is spread on top of the bottom geotextile.
The bentonite is covered with a water-soluble adhesive (to keep the bentonite in place), and the
cover geotextile is placed over bentonite. The stitch-bonding process is the next step for stitch-
bonded GCLs. In the stitch-bonding process, parallel rows of stitches are made connecting the
bottom and top geotextiles. The stitches are created in the machine direction the entire length of
the roll 100 mm between rows. The thread used in the stitch-bonding process is polypropylene.
Polypropylene is used because of its durability and its inert characteristics with respect to
chemicals. There is some flexibility in designing the reinforcing components of the stitch-
bonded GCLs. The stitch-bonding process can be changed by modifying the type of yarn, the
stitch bonding process, and the type of textiles used.
5.2.2 Internal Shear Strength
Mr. Fuller reported that the internal shear strength of Shearpro® includes a friction angle
of 5 degrees and a cohesion of 26 kPa, although the normal stress and shearing rate for which
these values apply were not specified. The cohesion is due to the interaction of the reinforcing
stitches and the fabric. The shear strength of the stitch-bonded GCLs is a function of the yarn
strength, fabric strength, and the stitch spacing. Therefore, the shear strength can be changed by
changing the yarn and fabric strength or stitch spacing.
5.2.3 Creep Shear Tests
Creep internal shear testing is being researched. Large-scale creep shear tests will be
performed at Purdue University. The speaker suggests that the reinforcing method used in
Shearpro® is not prone to creep. The reinforcing fibers in the GCL are not likely to "pull-out" of
the GCL when subjected to creep, because the stitch-bonding process ties the thread into knots in
the bottom geotextile. However, there is the potential for the reinforcing fibers to break or for
the polypropylene to creep. Interface shear testing of the GCL against different textured
82
-------
geomembranes will be pursued. The interface shear testing will help to see if the different
texturing processes (calendering, hot air, and HDPE spray) result in different shear strengths,
with the GCLs.
5.2.4 New Products
New stitch-bonded products are being created including a high tensile strength GCL that
has been used already in a California "canyon" landfill. An advantage of the stitch-bonding
process is its flexibility in using a wide variety of textiles. Also, the stitch-bonded GCL with a
nonwoven fabric provides a high interface shear strength. The interface shear strength can be
modified by altering the weights of the nonwoven fabric.
/ ; -v
5.3 Perspective of National S?al Co. with Respect to Bentofix®
Bentofix® is manufactured in North America by Albarrie Naue Ltd., and is distributed in
the United States by National Seal Company. The factory is located in Barrier, Ontario, Canada,
which is located near Toronto.
5.3.1 Manufacture .
The product is created on a continuous production line by dispersing dry sodium
bentonite onto a carrier geotextile at a minimum rate of approximately 5 kg/m2. The carrier
geotextile can be either woven or nonwoven (although the nonwoven is actually a
woven/nonwoven composite, with the woven side facing inward toward the bentonite and the
nonwoven component facing outward). The version of Bentofix® with a nonwoven geotextile
on one side and a woven on the other is Bentofix® NS, which is referred to as Bentofix II in
Chapter 3 and which was installed in the test plots with the woven geotextile facing downward.
The version of Bentofix® with a nonwoven geotextile on both sides is Bentofix® NW, which is
referred to as Bentofix® I in Chapter 3. A nonwoven geotextile is then placed over the bentonite
layer. The materials then pass through a series of needle boards where thousands of barbed
needles are punched through the composite. The needles carry fibers from the nonwoven upper
geotextile through the bentonite layer and entangle them in the carrier layer. The entangling
clumps are then thermally locked to the carrier layer by passing over a heat source. This process
encapsulates and stabilizes the bentonite .layer in the product. The, thermal-locking of the
needlepunched fibers to the geotextile gives the GCL internal strength and resistance to
deformation in a creep mode.
83
-------
5.3.2 Shear Tests
Short-term internal shear and long-term creep shear tests were performed to determine
the strength of the GCL and the results are included in the following paragraphs. Results of the
short term shear and long term creep tests are located in Appendix H.
The laboratory shear devices used in the creep and internal shear tests were large direct
shear boxes having dimensions 300 mm by 300 mm. The lower half of the shear box was larger
than the upper to allow for translation over a constant area. The normal load was applied to the
top by either an air bladder for large loads or by weights for lighter loads. The shear load was
applied by moving the lower half of the shear box for the short-term internal tests and by
applying a constant shear load to the lower box for the long-term creep tests.
Two internal shear tests series were performed on Bentofix® NS Thermal Lock. One
series of tests was performed at low normal stresses (34 kPa, 69 kPa, and 138 kPa), and the other
series at high normal stresses (965 kPa, 1170 kPa, and 1380 kPa). In each test the specimen was
placed in the test device dry, and then the target stress was applied. The test device was then
flooded, and the GCL was allowed to hydrated for 24 hours. After the hydration period, the
specimens were sheared at a displacement rate of 1 mm/min. The specimen area, in the tests at
higher normal stresses, was reduced to 200 mm by 200 mm in order to apply such a large stress
to the GCL.
When sheared under low normal stresses, the test specimens developed a peak shear
resistance during the first 12 mm of displacement, and then the shear stress decreased rapidly
after the internal connections were forced to fail. The peak apparent friction angle and cohesion
were 29 degrees and 33 kPa, respectively. The large-displacement apparent friction angle and
cohesion after 50 mm of displacement were 4 degrees and 6 kPa, respectively.
When sheared under high normal stresses, the specimens developed a peak shear
resistance, and then, with further deformation, the shear resistance dropped off as the specimen
was forced to fail internally. The peak friction angle under high stresses was 23 degrees, and the
peak cohesion was essentially zero. The large-displacement friction angle was 2 degrees after 50
mm of displacement, and the cohesion was 90 kPa.
5.3.3 Creep Tests
To evaluate how the GCL behaves during long-term shear (creep), two types of shear
tests were performed. The tests were performed to simulate conditions at the EPA field site, with
the normal and shearing stresses selected to be representative of the 2H:1V slopes. One type of
test was performed on both Bentofix® NS and Bentofix® NW at a normal stress of 20 kPa and a
shear stress of 10 kPa. The other type of test was performed on Bentofix® NS and Bentofix®
84
-------
NW at a normal stress of 413 kPa and a shear stress of 207 kPa. The shear devices for both types
of tests were 300 mm by 300 mm. In both types of tests the specimens were allowed to hydrate
and consolidate for 5 days prior to the application of the shearsjoad.
When the shear load was applied to the specimens under lower normal stresses, there was
an initial segment of movement. This adjustment to the application of the shear load rapidly fell
to effectively zero displacement versus time for both the NS and NW products. The specimens
were allowed to remain under the test conditions for nearly four thousand hours before the tests
were terminated.
In the long-term creep shear tests under higher normal stresses, there was an initial
segment of movement in response to the application of shear stress. The movement slowed and
the stresses were increased to a normal stress of 516 kPa and a shear stress of 241 kPa. There
was another initial response to the increased load, but it was smaller in magnitude than the first
response. As of August, 1995, the long-term creep shear tests at higher normal stresses are
currently at 500 hours of elapsed time.
After extrapolations of the results of these shear tests, the effective displacement rate was
determined to be 5 x 10'8 mm/min. It is believed that these date provide a good indication that
internal failure by creep is not a significant issue for the conditions examined so far.
5.4 Perspective nf fiSK Lining Technology with Respect to Gundseal®
Mr. Richard Erickson presented a brief description of Gundseal®, the GCL product
manufactured by GSE, and described two ongoing projects with GCLs. The two projects include
the GCL field test site in Cincinnati, Ohio, and a bentonite hydration monitoring test in the
Coffin Butte landfill in Corvallis, Oregon.
5.4.1 GCL Product
The Gundseal® GCL product consists of approximately 5 kg/m2 sodium bentonite with a
geomembrane backing. GSE manufactures several different polymer geomembranes that can be
used as the geomembrane portion of the GCL. The polymers include HDPE, ultraflex,
polypropylene, or any combination of the listed polymers given GSE's coextrusion process. The
geomembrane can be either smooth or textured, and its thickness can range from 0.3 mm to 2
mm. The length of the GCL product ranges from 46 m to 61 m depending on the thickness and
flexibility of the geomembrane. Gundseal® can be placed in the field with the bentonite facing
"up" or "down". Gundseal® is placed facing down in situations where the GCL replaces a CCL
in a composite liner. In these instances the bentonite is usually in contact with the subgrade.
The bentonite will hydrate, and therefore the strength of the bentonite becomes a concern in the
85
-------
design. More typically, especially in slope applications, is for Gundseal® to be placed where the
bentonite is facing upward. The geomembrane portion of the GCL is usually in contact with the
subgrade. A textured 0.76 mm to 2 mm geomembrane is used to increase the frictional
resistance between the subgrade and the geomembrane. In this case an additional geomembrane
is placed above the bentonite to create a "seal" so that the bentonite remains dry. The additional
geomembrane can be either a single or double-sided textured geomembrane.
5.4.2 GCL Field Test in Cincinnati. Ohio
The Gundseal® product has been installed in four of the plots at the GCL field test site in
Cincinnati, Ohio. Gundseal® has been installed in Plot A, F, and P with the bentonite side facing
up and covered by an additional geomembrane. Plot E was constructed with the bentonite facing
down and in contact with the subgrade. Plots F and P were constructed on the 2H:1V slopes
while plots A and E have been constructed on the 3H: 1V slopes.
The bentonite in Plot F is encased by two geomembranes and should have remained at
the manufactured water content. Instead, Plot F experienced wetting of the bentonite. Sampling
was performed to determine the extent of hydration of the bentonite within the plot. An area of
elevated hydration was found on the right panel of plot F near the entrance of the extensometer
cables. Also, the hydration seemed to "fan" out near the toe of the panel. It is not very clear
how the bentonite became hydrated, but it appears that the extensometer cables provided an
avenue for water to hydrate the bentonite. Because the causes of hydration were hard to identify,
GSE decided to construct another test plot, plot P, with the same configuration as plot F except
without the extensometer cables. Plot P was constructed in June, 1995, and the only
instrumentation placed in plot P was a 4 x 4 grid of moisture sensors. The moisture sensors in
Plot P indicate that the bentonite has not become hydrated.
At the same time that plot F was sampled to determine the water content of the bentonite,
plot A and E on the 3H:1V slopes were also sampled to make sure the moisture instruments were
functioning properly. The bentonite samples taken from plot A, which is encased by two
geomembranes, were still relatively dry. The one sample of bentonite taken from plot E, where
the bentonite is in direct contact with the subgrade, was at a water content of 50%. These water
contents were consistent with the moisture sensor readings.
5-4-3 Field Instrumentation of GCL/Bentonite Moisture Monitoring Program
The purpose of this project was to demonstrate the performance of Gundseal® with
respect to hydration and stability in a canyon landfill. The landfill is the Coffin Butte landfill in
86
-------
Corvallis, Oregon, and the instrumented cell is Cell 2C, which is a 5.6 ha cell with a double
composite bottom liner system. The lining system consists, from top to bottom,, of a 0.3 m
operations layer, a 0.3 m drainage layer, a 1.5 mm textured HDPE geomembrane, Gundseal^
GCL (1.0 mm textured HDPE geomembrane with the, bentonite side facing up), a 0.3 m leak
detection layer, a 1.5 mm textured HDPE geomembrane, and a compacted subgrade. The lining
system has an underdrain system because of a high water table.
The moisture instrumentation was installed for the long-term collection and processing of
moisture data for the bentonite within Gundseal®. Three distinct locations on the cell floor were
targeted: two locations with high potential for bentonite hydration, and one location for average
conditions. A total of eighty-eight fiberglass resistance block moisture sensors were installed.
They were placed in six transducer group arrays and in three pairs for redundancy. All moisture
sensors were covered with powdered bentonite so there would be no damage to the overlying
geomembrane. The moisture sensors were calibrated before installation. The data will be used
to extrapolate the behavior for the remainder of the cell.
87
-------
CHAPTER 6
REPORT OF ASTM D 35.04 SUBCOMMITTEE ACTIVITIES
Mr. Larry Well, the Chairman of ASTM Subcommittee D-35.04 on Geosynthetic Clay
Liners, provided a status report on standards development activities concerning GCLs with
ASTM. At the June, 1992, meeting of the ASTM D-35 Committee, a special session was held to
discuss the development of a standard practice for testing GCLs. A scope of work for a GCL
task group was developed. On the basis of discussions with Committee D-18 on Soil and Rock,
it was decided that Committee D-35 should manage the task group. During the June, 1993,
ASTM meetings, the task group was made into a new Subcommittee D-35.04.
As of last May there were 48 members on the subcommittee composed of 16 producers,
11 users, and 21 general interest members. The subcommittee is working closely with ASTM
Committee D 18.04 for Hydraulic Properties as related to the clay component of GCLs.
The new subcommittee comprises six task groups for physical properties, manufacturing
quality control/quality assurance (QA/QC), logistics, endurance, hydraulic properties, and
mechanical properties. Their task and activities are described as follows:
6.1 Physical Properties
Robert Mackey is the task group chairman leading the development of a standard test
method to determine the mass per unit area of the clay component of a GCL. The task group has
balloted a draft at subcommittee and main committee level. Comments received will be
incorporated to revise the draft for re-balloting. A round-robin testing program is being
developed to test the proposed method. The results of the ballot will be reviewed at the January
meeting and the standard moved forward.
6.2 Manufacturing OC/OA
Kent von Maubeuge is the task group chairman for developing a standard of practice for
manufacturing quality control and quality assurance of GCLs and a method for determining a
swell index value for clay and the fluid loss of clays. The first draft of the standard of practice
for QC/QA will be ready subcommittee ballot this fall. Draft ballots for standard method of test
for swell index value and for fluid loss of clays will be prepared if Committee D-18 agrees that
D-35 should do it. The results of the ballot will be reviewed at the January meeting and the
standard moved forward.
88
-------
6.3 Logistics
Robert Trauger leads as task group chairman in developing three draft standards. This
group is developing standards of practice for GCL installations, standards of practice for GCL
storage and handling, and standards of practice for GCL sampling. The first two items have had
one subcommittee ballot. Comments will be used to modify the draft for re-balloting at
subcommittee level. The scope of the procedure for GCL sampling was revised to remove
specimen preparation procedures and will be balloted this fall. The results of the ballot will be
reviewed at the January meeting and the standards moved forward.
6.4 Endurance
John Siebken is the task group chairman for developing index tests to evaluate the
chemical resistance or compatibility of the clay component of a GCL to liquids in soils. Tests
being evaluated for this use at this time include free-swell and fluid loss. These index tests may
serve to indicate further hydraulic conductivity testing is needed.
6.5 Hydraulic Properties
Scott Luettich is task group chairman. This task group is developing an index test for
measuring the flux or flow of liquids through GCLs. A subcommittee ballot was reviewed and
all issues were addressed and resolved. Plans are to co-ballot the re-draft at subcommittee level
with both D-35.04 GCLs and D18.04, and if possible get it to main committee ballot by October,
1995. The results to the ballot will be reviewed at the January meeting and the standard moved
forward.
6.6 Mechanical Properties
Alan Marr is the task group chairman for development of a standard for determining the
internal and interface shear strength of GCLs. At the task group meetings the details of the test
have been discussed and a draft of the standard will be balloted this fall. The results to the ballot
will be reviewed at the January meeting and the standard moved forward.
89
-------
CHAPTER?
REGULATORY STATUS OF GCLs
7.1 Perspective of U.S. EPA Superfund Headquarters
Mr. Kenneth Skahn discussed the Superfund program's perspectives concerning GCLs,
the reasons why GCLs are not incorporated into waste containment designs more frequently, and
how EPA documents might be modified to provide more information on GCLs and more
flexibility in landfill design.
The use of GCLs varies from region to region. The northeastern regions in the U.S. have
used GCLs in waste containment applications with success. However, the designs have been
very conservative; GCLs have been used on slopes less than 6H:1V. In Region 4, potentially
responsible parties (PRPs) have suggested the use of GCLs in landfills, but the communities are
skeptical about the equivalency between GCLs and compacted clay liners. The scattered use of
GCLs comes mostly from regulators' unwillingness to accept designs that differ from the
technical guidance document (U.S. EPA, 1989).
The EPA is involved in a research project with Sandia National Laboratory in which test
pads containing GCLs have been constructed (not on slopes) to monitor their performance with
respect to infiltration. One type of test pad has the configuration of a standard Resource
Conservation and Recovery Act (RCRA) Subtitle C cover, another test pad type is constructed
like the standard RCRA Subtitle D cover, and a third test pad has been constructed like the
standard RCRA Subtitle C cover except that a GCL has been incorporated in the design. In
phase one of the project, water will be sprayed on half of the test pads, while the other half will
be subjected to actual precipitation occurring at the site. The performance of the different test
pads will be compared. This project was constructed with the purpose of helping develop
designs for landfills in arid climates.
There has been some discussion with the RCRA program to amend the guidance
document (U.S. EPA, 1989) in order to provide alternative designs. These amendments are
considered desirable by the Superfund program.
7.2 Perspective of U.S. EPA Office of Solid Waste
Mr. Al Geswein discussed how three areas of EPA regulations and documents concerning
RCRA Subtitle D landfills interacted with GCLs. The three areas include the basic regulation
promulgated on October 9,1991, the guidance document issued in October, 1993 explaining how
EPA suggests states comply with regulations, and an upcoming EPA publication, or "issue
paper", which will provide information on GCLs.
90
-------
There are two places in the basic regulations where GCLs may be incorporated. These
include the bottom liner and final cover of landfills. A liner must be a composite liner described
in the regulations. If a designer desires to use an alternative liner instead of the prescribed
composite liner in a Subtitle D landfill, it must be demonstrated that the alternative liner will not
allow the concentration of certain chemicals to exceed specified concentrations at a point of
compliance, and the alternative must be approved by the Direction of an approved state.
Therefore, there is no restriction to the use of GCLs, and a GCL may be chosen for design as a
component in the liner, provided the required demonstration concerning impact to ground water
is made and the Director of an approved state concurs with the use of the proposed alternative.
A final cover in a Subtitle D landfill is required to not let more fluid into the top of the
landfill than the amount which exits the bottom of the landfill. There is no restriction on the type
of materials; therefore, a GCL may be used in a final cover. There is no formal restriction of the
use of GCLs in the regulations.
The second area which EPA Office of Solid Waste is involved with GCLs is the
supporting technical manual issued in 1993 (U.S. EPA, 1993). The technical manual includes
three sections: a regulations section, an applicability section stating who must comply with the
regulations, and a technical section which addresses technical concerns and gives suggestions on
how to comply with regulations. GCLs are mentioned once in the technical portion of the
guidance document. At the time of development of the technical document there was little
published information about GCLs, and therefore there was little included in the technical
guidance document.
The third area where the EPA's Office of Solid Waste interacts with GCLs is in an
upcoming publication called a fact sheet. The fact sheet is a one-page summary which will
discuss how GCLs might be used in Subtitle D landfills. The fact sheet will describe the
technology, summarize case studies where GCLs have been used, discuss how GCLs fit into the
regulations, and provide a list of references on GCLs.
91
-------
CHAPTER 8
PANEL DISCUSSION
8.1 Critical Issues Concerning GCLs - EPA Project Team
8.1.1 Intimate Hydraulic Contact vs. Shear Strength
David Daniel pointed out that, based on the slides that have occurred at the EPA test site
where sliding occurred at the interface between a woven geotextile and a textured geomembrane,
and based on the discussion during the conference, designers might consider placing the thick
nonwoven geotextile of a GCL next to a textured geomembrane. However, this would reduce
the quality of the hydraulic contact between the geomembrane and GCL, since a thick,
nonwoven geotextile would separate the bentonite from the geomembrane. Placing a thick,
nonwoven geotextile next to a geomembrane on steep slopes may not matter as much as on
flatter slopes. Designers should be cautioned to not forget about hydraulic properties of GCLs in
efforts to increase the shear strength of liner systems and interfaces.
8.1.2 Internal Shear Strength of GCLs under High Normal Stresses
Rudy Bonaparte discussed an issue of growing concern, especially in the design of
canyon landfills. The issue is the internal shear strength of GCLs under high normal stresses and
long-term loading conditions. There is not much available in the literature that describes the
limits of GCLs under these conditions. All that is available is the knowledge that mostly all of
the GCL products have large-displacement shear strengths of about 3 to 5 degrees under very
high normal stresses. More information about the behavior of GCLs at high normal stresses
must be obtained. The GCL manufacturers are challenged to continue their effort to produce
new or modified GCLs with higher internal shear strengths and decreased strain-softening
behavior under high normal stresses. Possibly .this can be accomplished by using different types
of bentonite, adding granular admixtures such as sand in GCLs, or through other methods.
8.1.3 Long Term Effects
Robert Koerner touched on the issue of long-term effects. Since GCLs have the potential
to be included in many long-term projects, the ability of GCLs to perform after long periods of
time should be determined. One factor that seems to limiting the more widespread use of GCLs
is the uncertainty of its performance in the long term. Long-term tests are not performed often
because of the difficulty in testing.
92
-------
8.2 Questions by the Audience
8.2.1 Do Bid Projects vs. Designs Need to he Sole-Sourced?
David Daniel commented that construction specifications can be developed to allow for
alternatives so that bid projects and. designs do not need to be sole-sourced. Rudy Bonaparte
noted that if a designer desires to include a GCL as a component in a waste containment facility
he or she must, for example, perform project specific testing to establish the shear strength. The
designer has an option to proceed in two different ways. The first option is to test all of the
products and prequalify them all during the design stage. The second option is to limit the
testing program to determine the validity of the design, then write the specifications that require
the successful contractor, as part of initial activities, to carry out further testing. The additional
testing will prove that the specific product the contractor is going to buy and install for the
project will be stable in the design. The second option eliminates the sole-sourced problem.
8.2.2 What is the Suitability of Replacing a Compacted Clav Liner CCCU in a Composite Liner
with a GCL?
Rudy Bonaparte commented that most research indicates that substitution for a CCL with
a GCL would be favorable, but the substitution must be made on a project-specific basis. The
ultimate goal in constructing a liner is to protect ground water, and for double liners, a GCL in
the primary liner would be an adequate component in the attempt to achieve that goal.
Robert Koerner remarked that this topic has been the subject of several papers authored
by Koerner and Daniel. In the discussion of technical equivalency between a CCL and GCL
there are 25 issues. Some of these issues include hydraulic properties, shear strength,
constructability, and performance. Only in the issues of adsorption and subgrade conditions is
the GCL not equivalent to a CCL. However, these two issues can be bypassed with proper
construction quality control and assurance.
David Daniel offered several observations. To begin, Subtitle D regulations do not allow
for a direct comparison between CCLs and GCLs. The regulations do not require it to be shown
that a GCL is superior to a CCL when replacing a CCL with a GCL. If a CCL is to be replaced,
it must be demonstrated that the replacement material will not allow unacceptable impact to
ground water at a point of compliance. It is curious that the regulations do not require that a
CCL be demonstrated as having no impact to ground water with a prescriptive design. It is Mr.
Daniel's belief that if all of the 0.6-m-thi.ck CCLs in Subtitle D landfills in the U.S. were
replaced with GCLs, and assuming all GCLs were stable on the slopes, the GCLs would
generally outperform the CCLs. This is mostly a result of the incorrect construction of CCLs
and the desiccation/cracking of CCLs. As smaller landfills with less technical capability for
93
-------
building and construction, the advantages to GCLs increase, because it is easier to install a GCL
than compacted clay. The largest concern in the discussion of equivalency is the resistance to
puncture. A GCL is never going to be equivalent to a thick layer of compacted clay in regards to
resistance to puncture. Therefore, it is a fair and appropriate question to ask what will happen if
the GCL is punctured, especially in a hydrogeologically sensitive area.
8.2.3 Should Designs for Waste Containment Structures Be Based on Peak or Residual Shear
Strengths?
Rudy Bonaparte commented that designs should be project-specific. More generally,
designs may be based on slope stability evaluations using peak strength, while checking the
design for stability using residual strengths. David Daniel noted that if the factor of safety
calculated for slope stability using the residual strength is greater than one, then the design is
generally acceptable. If the factor of safety with residual strengths is less than one, then it must
be determined if residual conditions will develop and how critical those conditions will be.
Robert Koerner noted that the design should include residual strengths because of the long-term
question.
94
-------
CHAPTER 9
REFERENCES
Amble, T. (1994), "In-Plane Hydraulic Conductivity and Rate of Wetting of Geosynthetic Clay
Liners," M.S. Thesis, University of Texas at Austin, 136 p.
Boardman, B.T. (1993), "The Potential Use of Geosynthetic Clay Liners as Final Covers in Arid
Regions," M.S. Thesis, University of Texas at Austin,109 p.
Bramlett, J.A., Furman, C. Johnson, A., Ellis, W.D., Nelson, H., and Vick, W.H. (1987),
"Composition of Leachates from Actual Hazardous Waste Sites," United States
Environmental Protection Agency, Cincinnati, Ohio, 113 p.
Daniel, D.E., and Estornell, P.M. (1991), "Compilation of Information on Alternative Barriers
for Liner and Cover Systems," U.S. Environmental Protection Agency, Risk Reduction
Engineering Laboratory, Cincinnati, Ohio, EPA 600/2-91/002.
Daniel, D.E., and Boardman, B.T. (1993), "Report of Workshop on Geosynthetic Clay Liners,"
U.S. Environmental Protection Agency, Risk Reduction Engineering Laboratory, Cincinnati,
Ohio, EPA 600/2-93/171.
Daniel, D.E., Shan, H.Y., and Anderson, J.D. (1993), "Effects of Partial Wetting on the
Performance of the Bentonite Component of a Geosynthetic Clay Liner," Geosynthetics '93,
Industrial Fabrics Association International, St. Paul, Minnesota, 3: 1483-1496.
Erickson, A.E., Chamberlain, E.J., and Benson, C.H. (1994), "Effects of Frost Action on Covers
and Liners Constructed in Cold Environments," Proceedings, Seventeenth Madison Waste
Conference, University of Wisconsin, Madison, Wisconsin, 198-220.
Harpur, W. A., R. F. Wilson-Fahmy, and Koerner, R.M. (1993), "Evaluation of the Contact
Between Geosynthetic Clay Liners and Geomembranes in Terms of Transmissivity,"
Proceedings of the 7th GRI Seminar, Geosynthetic Research Institute, Drexel University,
Philadelphia, 138-149.
Hewitt, R.D. (1994), "Hydraulic Conductivity of Geosynthetic Clay Liners Subjected to
Freeze/Thaw," M.S. Thesis, University of Texas at Austin, 103 p.
LaGatta, M.D. (1992), "Hydraulic Conductivity Tests on Geosynthetic Clay Liners Subjected to
Differential Settlement," M.S. Thesis, University of Texas at Austin, 120 p.
McNabb, G.D., Payne, J.R., Harkins, P.C., Ellis, W.D., and Bramlett, J.A. (1987), "Composition
of Leachates From Actual Hazardous Waste Sites," Land Disposal, Remedial Action,
Incineration and Treatment of Hazardous Waste, Proceedings of the Thirteenth Annual
Research Symposium at Cincinnati, Ohio, EPA/600/9-87/015,130-138.
95
-------
Ruhl, J.L. (1994), "Effects of Leachates on the Hydraulic Conductivity of Geosynthetic Clay
Liners (GCLs)," M.S. thesis, University of Texas at Austin, 101 p.
Sai, J.O., and Anderson, D.C. (1991), " Long-Term Effect of an Aqueous Landfill Leachate on
the Permeability of a Compacted Clay Liner," Hazardous Waste & Hazardous Materials,
8(4): 303-312.
Stanforth, R., Ham, R. Anderson, M., and Stegmann, R. (1979), "Development of a Synthetic
Municipal Leachate," Journal of the Water Pollution Control Federation, 51(7): 1965-1975.
Texas Water Commission (1985), Technical Guideline No. 1, "Waste Evaluation/Classification,"
Austin, Texas.
U.S. Environmental Protection Agency (1989), "Technical Guidance Document: Final Covers on
Hazardous Waste Landfills and Surface Impoundments," EPA/530-SW-89-047, Washington,
D.C., 39 p.
U.S. Environmental Protection Agench (1993), "Solid Waste Disposal Facility Criteria,
Technical Manual," EPA 530-R-93-017, Washington, D.C., 349 p.
96
-------
Appendix A
List of Attendees
-------
-------
a,
O
K
S
g
• ^ fc- M ^
•U 00 Q ffi <,
r-
w 00
•S* =S§
lHa*5
5 c »- g?
C -£3 >n J?
S3
"S
S!l
•1^
U^I
!fl«
§J<-1
!lil
J is
•?; «
y in
3 < «
? o i
o.>£
s.o =
£ £ S o 2
^ II
CO ^ 3
^^^
^ £ W
So Z
t-r U-l 10
O ro (S
Tf —« OO
g
s
II
.s
g-
O^D
c 8
S ffi
•^ K M
a £
H«5
•S 2
>!!<
H r~<
w a
o u
A-l
-------
I ss
"sit
sfslg
Sill
A-2
-------
to
.2 S .2 j? =
o i2 o a a,E
£*<
* &
.a .2
•§ 3
< ** I "I1 ^ w
^=«^§« °s
foaS-S &|.s.s^
a- o o
[ 5 O O
A-3
-------
1
.S2
8
S
Infrastructure
«8
Rust Environmenl
U
o
t4->
o
!
S3
1
Univ. of Western
Kenvirons, Inc.
ce Conservation Con
1
£
."3
!•£
3 l
x S
£t
12
g
. w
>
u < [§
a S s
3 0 g
rsl
M
1
tn
k
1 w
S .2
m o
s
V)
£
fli
Tensar Environmi
i
1
<*-
o
s.
!l
™ O
3i
i
M
ll
•s-s
ai
1J
J §?
£sS
(A
i*
11
Browning-Ferris ;
National Seal Coi
1
New Mexico Env:
O
&M
U^S
S B' c
4li
s
S
^1|^4
"I « Q *. S fe S & * e S d
|S | I nl 1-g^ll s|
5 g>§ g g 5>J='r:£55£!°--
Ctf^BiBioicdoSc^^incoKicnmc/ioo
|<
•o ^>
< g g- u
r .1 ,<3 •«
1 W NJ H g
D
OS
U
.<^
is
j«4
« d.
S 'i a • •
«> S "~ .?
c -p 3 -2
^ a2
^•g
*!
3 s
CO t/3
A-4
-------
I
A-5
-------
-------
Appendix B
Total Down-Slope Displacement in Test Plots
-------
-------
:
1
?
: o
• i ^
; ~ ^
o
.1
5 w
•X *° o o e
•> ' o in c
fc 3 "* *~ T
|l ' ' c
if
5 OD
f!
1«
«|
a
3"
!
i s ^.
\ *- o ,
§0 c
m c
CM ^ T-
g •=•
o - w
« g"
2.
i O
•z p
•I
0 • -
0 "^
w m V
•c
•^
s!
IX
_//
^Jf-
-y'"
~~
3 O O O <
3 m o m <
- o o o i
BSHOUj) JUBUJOOB|dS!
O "Sf
«. (0
S -|
c
f
^p
r»
"^
5*
^^
"^^
^>
//
I—I
—
—
> O O 0 C
> in o in e
O O O i-
i
"S •
o>
K. '•
,
, i . g
Y i °
'S v-
1 — 55. ^S ~
Jin «^>
o
D)
-- : ^
11
a:
2"
- ,,.g
' I v
if » ^
;r;,i» i
_i ! ' o
i ; I
• :
i ' '•
in o a . : o m c
*- CM I '. CM T- T-
0 <-*
0 T S
n S"
i ^
; «T . o •
E g
?k p -
\_i £
j-^^Tp
•P-^
nl-: =
^T
^^ "p
|? £
T/ 1
E5 •
— 0)
s-
—
3 O O O O O O2
3 in o in o m oQ
- o d o »- i- co
BEiqoui) )uauie3e|ds|Q
n
>>
o •§
o - —
m e
P »
C
CO
~ m Q-
= • i
^ E
^ • _
h
r-^-4! , _
_' o
r^' • *
S\* c
7 2
c • .
^
•i
^- — — 1 c
9OOOOOOO
>inoinomoo
d ci d *- »- cj
JU9UJ8OB|dSjQ
B-l
-------
o
o -
co
•§, I I
2 i I
O
o
i! v+-
in
g>
o
id
o
q
pa W ^
O
in
O
in
o
o
(«BlpU|) >U8UJBOB|dS|a
o
o
d
o
o
JU8tU93B|dS|a
00
o O
g -
p
1 -
*
i I
o
S §
o
o
O
m
O
o
(••tpu|)
I '
S
! i
8
0
o
O '
in
?
o
o
o
o
o
(sai(3U|) |uauiaoe|ds|Q
o
in
o
o
B-2
-------
o
O
! 1
i 2
2
CO
Q.
•a
ai 000000000° . oooooooooo
•> s> oinoinoinom.oO omoinomoinoo
t o" *!* T T
-ooO'-r-ej cg»-»-ooo^-'-co
|S (SOMOUI) tuemaoeidsia (SOMOUI) jueuiaoeidsia
8?
3 S
ft!
||
1 ;
0 "Sf i ° -ST
o - 2. 1 o - S '
co £• I 1 co g1
ti i : ' § i
^ ; P . s : - j ^: P s
/ 1 s> .g>
0
(oogoui) )uouiaoe|ds|Q
oo
in o
in o in o o
(•eipui) |U8iueoe|dsia
B-3
-------
8
s
' &
§
0.
1
01
§•
§o o
in o
o
in
o o o
in o in
d T-' W
' ' S>
s.
0
o
0
in
0
o
0
in
00
o O
CM T- T-
juauiooB|dsia
_i
• 8
o
I
o>
E
o
C3
<
in
§.
o
o
O
o
V T 9
o o o o
p in o in
cj i-' i^ d
o
in
in
d W W
B4
-------
' O
i O
o
m
(9
o
s.
I
o
O
CD
I
cc
1
fw
ci
W g !
H 1
||
DY-STATE DISPtA
jndseal - Bentonlte
i
< 5
w
to lii
ST
f ' '."li-
I 1 • i
o in 010 o mo u> OQ
CM»-»-OOO'-'-eM
(seijoui) }ueuiQOB|dsia
If §-
S.
i* i
H 5 • ^ 1 • '
^ 1 1 "
: 1 ^
oooooooooo.
O IO O IOO to O IO O Q
Ni-T-OOO»-»-CM
(8eMOU!) »u8UJe8BIdsla
! i • I '
JB,
B 1 C
• z i a
»~ 1 2
V f
. (9
^ I
IO p IO O IO O IO
Wi-^oooi^i^cvi
B-5
-------
o
o
S '
o .
o
o
o
o
o
o
o I
§o
qi
co ^--
i i
'. ^" 6
|l "
1
P|
5" o
O I CM
1 1
in
i I
3 °
1
1
i
§ § i
CM T^ C
(•01)3
I ..
1
| B C '; | '
1 2
s>
1 ; K
'.-. '
1
• I
' «
1
£
3 O O O O O
> W CM CO ^f
u|) )uau)aoe|dsia
• S
^, o
p co
I
o
CM "
O
o
o
CM
s
O O O O
q q q o
CM ^ d t-:
ii
£ i
I
i
S
CM
)uauia9i|d*ia
0 o
0 0
CM CO V
o i
o
0
CM T-
o o
00
o
0
0 »^
o o o
o o q
CM CO It
)U*UI03(|df|a
B-8
-------
1
O_T
«=•:
e
CD
o
O
B i
1 I
£
o
o
cj W o
o
o
o
o
o
o
o
o
o o
O O
in
i
2
a
O
o
o
m
o
o
CM
O
o
a.
1
tL
1
i §
Trc
^
o
p ,
i
1
0
p •
o o o • o
o o p • p
(69l|OU|) JUaU19DB|dS!Q
1 fe
1 7
000
pop
e\i ^ o
1
1
§ §
i- CU
O 0
o o
CO W
o
p
in
(seqouj) )ueuiooe|ds|Q
3
O
(9
I
§
o o o o
o o o o
o
p O
in
)ueuieoe|dsta
B-9
-------
g ,
S g 8
o
o
o
o
o
in
e> i-' -p-'
S
10
B
en
i
o o o
p in o
t\i T-' T-'
g
o
to
£
_D> ' i
DC i
I
O O
o in
§<
juoui90B|ds|a
-^ o
» o -
>> CJ
S
OJ
o
10
§ g
o o
o in
o
o o o
o in o i
e T-' T^ oi
}U8UJ»3B|d8|a
O
(9
I
in
•
o>
I,
ir
S.
3
o
o
CVI
0
\
o
o
o
in
9
§
o
o
in
o
0
o
^
g
•t-
o
CVJ
B-10
-------
o
O
0>
D>
IB
O
SSli
PT:
o
s
61
o
0
c\i
§1 ii
OQOOOOUUO
in p in o in o in po
^ -r^ p d d *-: T-' c\i
(S9L|OU|) }U9UJ8Oe|dt:!a
o
tC" O -
>> w
F
0 <=
o
o
!
0
a>
i
3 O
s u>
>i ^
1
1
1
1
0
p
>
/
r-f
°.
d
^\
A /
?
o in
d d
1
1
1
i
o
p
o e
10 t
^ e
S.L
o
CJ
(I g 1
I
8
-T
s,
1 \i
S
r *=v
OOOOOOO Oo
moinoinoin OQ
B-ll
-------
8
cvi-r-,-oo6~-^c\i
(seuoui) juoiU8oe|ds|a
5!
I.-. °
i
! -s
I I
! |
' 1
d
o
i
i
i
£
gg. gggggSSfi
O 10 O Q
juauiooeidsia
o
O
1
O
(9
I
O
8 g §
QJ T» T-
S I H-
o
O O O O O
in p u> o in
>U|) )UaUIB3B|dS|a
tn '
O o i
po !
N 1
!
• •• i
o o o o o
o u> o in o
' * T 7 9 °
(«9MOU|) JU9U
1 S
• I
O O O O S I
mow o,§
O i— -^ CNJ
ieoB|ds|Q
B-12
-------
if
!
o
o
n
t
ff
^-
I
*
I
o in o 10 o 10 o u> of
oirrPodW'^ci
I
1 ! i
£ ! i
e i •
(9
o
o -
-------
PLOTP
Gundseal (bentonile up) • 2:1 Slope
100 -
I a, 90 •
£ 80
: 1 70
' & 60
o 50
i $ 40
: * 30
' £ 20
Z 10
' 0
Moisture Reading
-*rOU.ttaia>vja>COO
0
100
a 90 -
= 80 •
1 70 •
£ 60 •
o 50 •
3 40 •
S 30 •
£ 20 •
10 ' _=
0 — 0-
0
I 100 •
_ 90 -
f BO-
•g 70 •
£ 60 -
o 50 •
J 1 40 •
S 30 -
i £ 20 -
• 10 ' _^
0 o
0
• Inner left
panel
Moisture Readings vs. Time
Plot Location: Crest
— •— p m
10 20 30 40 50
Time (days)
Plot Location: Upper Middle
=f — i- • '; •'.„
10 20 30 40 50
Time (days)
Plot Location: Lower Middle
10 20 30 40 50
Time (days)
Plot Location: Toe.
10 20 30 40 50
Time (days)
= outer left * Inner right 0 OU|er right
panel panel panel
B-14
-------
Appendix C
Differential Displacement in Test Plots
-------
-------
PLOT A
Displacement Above GCL - Displacement Below QCL
Gundseal - Bentonite Side Up-3:1 Slope
Relative Displacement vs. Time
Left Panel
I -3.00 -
§ -2.50 -
"5.
5 -2.00 - A
I
£.£ -1.50 - .
j J
go -1.00 - ^
§ I
5 § -°-50 - v
(/
Time (days)
200 -' 250
Extensiometer No.
Relative Displacement vs. Time
Right Panel
Extensiometer No.
C-l
-------
. Displacement Above GCL - Displacement Below GCL
PLOT B ' Bentomat - 3:1 Slope
f I -2.00 -
i £
I -1-50-
Q -1.00 -
Relative Displacement vs. Time
Left Panel
Time (days)
200 250
Extensiometer No.
jT.§ -0.50 -
J J °
§ O 0.00 f
Relative Displacement vs. Time
Right Panel
Time (days)
Extensiometer No.
C-2
-------
Displacement Above GCL - Displacement Below GCL
PLOT C Claymax - 3:1 Slope
i
1
i
i
I
! i
i
Q.
W
a
t
XI
JS /5
si
P
I
01
g
: s-
5
i
: o
ai
o
CD
8-
i Q
i i
— ' !§.
i *7 -i
og
o o
: s
p
! 8
i S
Relative Displacement vs. Time '
Left Panel " •
-2.00 - :
-1.50 - ;
i
-1.00 T
.o.so I ^ ^ ^^^^^^ :
o oo ^B-^^iv^^^^^i1^ ^^i*^^""* ^^^i— i«>o
0.50 7 /f Time (days)
1.00 -
1.50 - ^
2 00 —
Extensiometer No.
r-^
• I ' " 2 * 3 -O 4 A- 5
Relative Displacement vs. Time
Right Panel
-2.00 -
-1.50 -
o ' ' •
-1.00 - , *
• • '• ' '
-0-50 - 5()^ ^ 100 J ^V wt4t^Tt^9^ ^~t^
Q _ ir~i>-V/^^"^--4 r^~*~~*~S^'» P*~--^^Ng"--— --*' B — a "
\Jt\J\J HU • (ft B'^^C 5! "^^ H.1^^" ^^ffV? *5 A "**? A A A A A i
• • "™ !** ^^^S • iSit A A A AA AA
0.50- \ / 4 Time (days)
' ; V
1.00 - • »
1
1.50T vtf i
2.00 -1
Extensiometer No.
-* 5 '
C-3
-------
Displacement Above GCL - Displacement Below GCL
PLOTD Bentofix I! (NW up) - 3:1 Slope
-3-00 -
-2.00 -
Relative Displacement vs. Time
Left Panel
/-
Jm-m-mm—m-m-m—m-m- •-
i en n*c&
Time (days)
Extensiometer No.
3.00 -
Relative Displacement vs. Time
Right Panel
me (days)
Extensiometer No.
5 i
C-4
-------
Displacement Above GCL - Displacement Below GCL
PLOTE Gundseal - Bentpnite Down - 3:1 Slope
I
-1-50 -
-1.00 -
o <5
8<3
Be
Displa
Relative Displacement vs. Time
Left Panel
-»—»
• A i A
>• \ * / '•.
Extensiometer No.
Time (days)
Relative Displacement vs. Time
Right Panel
Time (days)
Extensiometer No.
C-5
-------
Displacement Above GCL - Displacement Below GCL
PLOTF Gundseal - Bentonite Up - 2:1 Slope
-3.00 -
1 I
Relative Displacement vs. Time
Left Panel X
Time (days)
0 50 100 150 )
^_Ti^''*"gsE==gg'fta'"=^^'=^^^ Mjri
200 25
5.00
8 0
a> S 7.00
> £
| 0 9.00
| 11.00
1 13.00
15.00
Extensiometer No.
en
ii
-ici}
bb
oo
'0 '
f
Q
I
11
, $$
c S
> £
2 e>
< CD
! C
e> 1
E
i o>
• ° •(
o 1
I o.
\ &
i 1
1.00 -
14
3.00 -
5.00 -
7.00 -
9.00 -
11.00 -
13.00 -
IS. 00 -
Relative Displacement vs. Time
Right Panel
Time (days)
50
100
150 J
___ —_-_-V=j".
» « » »« « »«—«•
200
250
Extensiometer No.
C-6
-------
Displacement Above GCL - Displacement Below GCL
PLOT G ' Bentomat- 2:1 Slope
Relative Displacement vs. Time
Left Panel
g -2-00 -
1 -1-50 -
a-
3 -1.00 -
£.£ -0.50
Time (days)
100
150
200
Failure
Extensiometer No.
Relative Displacement vs. Time
Right Panel
g -2.00 -
Displacement Above GCL (in.) - Displace
Below GCL (in.)
-1.50 -
Time (days)
-1.00 -
-0.50 - A
0 ; /*~\ 50 100 150 200
0 00 !!• •§ / * * jjrrr"ll
u "T ~7~-=-c^^5
0.50 - J*/
1.00 - <>KS>-^>
1.50 - Failure
Extensiometer No.
• — •
• | —• •_• g * 3 "C>" 4 — * 5
C-7
-------
Displacement Above GCL - Displacement Below GCL
PLOT H Claymax - 2:1 Slope
Displacement Above GCL (In.) - Displacement
Below GCL (in.)
Relative Displacement vs. Time
-2.00 - Time (days) .
-1.009 5 10 15 20 25 30
.00 * B^*- ^ '
1.00 - \
2.00 | \ !
3.00 + **\ i^-"4! '
4.00 \ \
5.00 - \ \
6.00 - \\
7.00- \\
8.00 - \'\ Failure
9.00 - ' \
10-°° - ^Extensiometer No.
•^ A
Displacement Above GCL (In.) - Displacement
Below GCL (In.)
Relative Displacement's. Time
-2.00 - Time (days)
-1.00 0 5 10 15 20 25 30
1.00 -
2.00 - m-»
3.00 - jr^~^*^V,
4.00 - / \ \
5.00- Nf \\
6.00 - \\
7.00- \\. ™ure
8.00 - U i
9.00 - l\
10-°° - 'Extensiometer No.
• 1 2 • 3 0 4 A 5
C-8
-------
Displacement Above GCL - Displacement Below GCL
' Bentofix I - 2:1 Slope
-4.00 T
I 1.00 -
u
£
B 2.00 -
Relative Displacement vs. Time
Left Panel
Extensiometer No.
-* 5 •
-4.00 -
Relative Displacement vs. Time
Right Panel
00 •*—X* 260
2.00 -
Time (days)
Extensiometer No.
C-9
-------
3.00 -
Displacement Above GCL - Displacement Below GCL
PLOTJ Bentomat - Granular Drainage - 2:1 Slope
Relative Displacement vs. Time
Left Panel
Time (days)
200
250
Extensiometer No.
Relative Displacement vs. Time
Right Panel
to—•-•-•—m-e •—•
I 2.00 -
3.00 -
Time (days)
Extensiometer No.
-• 1
-• 3
C-10
-------
Displacement Above GCL - Displacement Below GCL
PLOTK Claymax - Granular Drainage - 2:1 Slope
Relative Displacement vs. Time
Left Panel
-3.00 -
Time (days)
Extensiometer No.
* *—»
-3.00 -
Relative Displacement vs. Time
Right Panel
Time (days)
250
Extensiometer No.
C-ll
-------
Displacement Above GCL - Displacement Below GCL
PLOT L Bentofix I - Granular Drainage - 2:1 Slope
Displacement Above GCL (In.) - Displacement
Below GCL (in.)
Relative Displacement vs. Time
Left Panel
-3.00 -
Time (days)
-2.00 -
-1.00 -• , M^ 1
0 m 50 x^Ocx^PO 150 J ^OO 250
o.oo *» &£^£O-^E^ ^'^^^^^^^aa.^g^^^jr^S^aL.B^^-S — i
1.00 - •' ,
2.00 -
4.00 - A
5>0° " Extensiometer No.
Displacement Above GCL (in.) - Displacement
Below GCL (in.)
Relative Displacement vs. Time
Right Panel
-3.00 -
-2.00 - Ctf( Time (days)
-1.00 - J '
0 50 100 150 1 200 ^2j£0
A f\n w • ^^k~A — ~"A ^'— -=s~~:~--~jf*^~~~- Jy^^ ~*^Srf • • >^\ * M
' T / *w
1.00 T* / >
2.00 -i
3.00 T
s-°° ~ Extensiometer No. [
C-12
-------
-2.00 -
Displacement Above GCL - Displacement Below GCL
PLOTN Bentofix II (NW up) - 2:1 Slope
Relative Displacement vs. Time
Left Panel
Time (days)
Extensiometer No.
2 « 3
4 * 5
2.00 -
Relative Displacement vs. Time
Right Panel
Time (days)
200 250
Extensiometer No.
2 * 3
C-13
-------
-------
Appendix D
Moisture Instrument Readings in Test Plots
-------
-------
PLOT A
Gundseal - Bentonite up - 3:1 Slope PLOT A
Moisture Readings vs. Time
Plot Location: Crest
100 150
Time (days)
200
250
0 -'
Plot Location: Middle
100 150
Time (days)
200
250
Plot Location: Toe
50
Subsoil-GYPSUM
100 150
Time (days)
200
250
GCL/Subsoil -
FIBERGLASS
w/in GCL -
RBERGLASS
D-l
-------
PLOTB
Bentomat - 3:1 Slope
Moisture Readings vs. Time
Plot Location: Crest
50
100 150
Time (days)
200
250
100 -
90 -
g 80 -
1 70-
« 60-
0 50 -
3 40 -
Plot Location: Middle
I 30- -k
g 20- \
10- *M-»
0 —*
50
100 150
Time (days)
200
250
0 —A A A A A-
0 50
Plot Location: Toe
' A A A A
100 150
200
250
Time (days)
Subsoil-GYPSUM A- GCL/Subsoil - FIBERGLASS
D-2
-------
PLOTC
Claymax-3:1 Slope
100 -
90 -
§? 80 -
\ ^ 70 i
! * 60 *
: g 50-
3 40 -
•5 30 -
S 20 7
10 T
_ •• M
0 r~
0
1 .
Moisture Readings vs. Time
Plot Location: Crest
A-4^— A |*| ^
' ^-t-^L-^-^-A+t^M • • • ,-
^•nlnfe B*"^ • '
// '' ' r
/ \ / • • v/> ' ;
/ \ / .' / *
/ < ' '
/ V' . - . ' ' /' '
/ *
50 100 150 200 250
Time (days)
100 -
90 -
W 80 -
1 70 '
« 60-
o> 50 -
3 40 -
I 30-
£ 20 -
10 -
0
Plot Location: Middle
4
1
• !••§• i_ ,_.
50 100 150 200 ;*. 250
Time (days)
100 -
90 -
o> 80 -
S 60 -
c
o> 50 -
• 3 40 -
! •§ 30-
S 20 -
10 -
_ -
! 0
Plot Location: Toe
..'• •
\
X
' /•"*'' *
50 100 150 200 250
Time (days)
• o..u.«*.n _ /^VDOt AH A tZC-.l ACnhcAil . PIRPnm AQft
D-3
-------
PLOTD
Bentofix II (NW up) - 3:1 Slope
Moisture Readings vs. Time
Plot Location: Crest
100 150
Time (days)
200
250
Plot Location: Middle
100 150
Time (days)
200
250
Plot Location: Toe
100 150
Time (days)
Subsoil-GYPSUM
200
GCLYSubsoil • FIBERGLASS
250
D-4
-------
PLOTE
Gundseal - Bentonite Side Down - 3:1 Slope
Moisture Readings vs. Time
Plot Location: Crest
100 150
Time (days)
200
250
Pl°t Location: Middle
100 150
Time (days)
200
250
Plot Location: Toe
100 150
Time (days)
Subsoil-GYPSUM
200
GCL/Subsoil - FIBERGLASS
250
D-5
-------
PLOTF
Gundseal - Bentonite up - 2:1 Slope
Moisture Readings vs. Time
Plot Location: Crest
100 150
Time (days)
200
250
Plot Location: Middle
100 150
Time (days)
200
250
100 150
Time (days)
200
250
Subsoil-GYPSUM
GCL/Subsoil •
FIBERGLASS
w/in GCL -
FIBERGLASS
D-6
-------
PLOTG
Bentomat - 2:1 Slope
D>
'o
i
0)
100.0 -
90.0 r
80.0 :
70.0 :
60.0 :
50.0 -
40.0 *
30.0 T
20.0 '
10.0 -
0.0-
Moisture Readings vs. Time
Plot Location: Crest
50
100
Time (days)
150
200
100 T
90 -
i> 80 -
H5 70 -
g 60 -
£ 50 -
•5
Plot Location: Middle
40 -
30 -
20 i
10 -
n —
• X
wf
A
— -A.
— A A A A - -.-.
50
100
Time (days)
150
200
100 -
90 -
o> 80 -
:• ^
50
Plot Location: Toe
100
Time (days)
Subsoil - GYPSUM
150
GCUSubsoil - FIBERGLASS
200
D-7
-------
PLOTH
Claymax - 2:1 Slope
Moisture Readings vs. Time
Plot Location: Crest
50
100
Time (days)
150
200
100 -
90 -
SO-
70 "
60-
50 "
40 -
30 -
20 -
10 -
o —
Plot Location: Middle
50
100
Time (days)
150
200
Plot Location: Toe
50
Subsoil-GYPSUM
100
Time (days)
150
GO/Subsoil- FIBERGLASS
200
D-8
-------
PLOT I
Bentofix I -2:1 Slope
100 -
90 -
? 80 -
Moisture Readings vs. Time
Plot Location: Crest
100 150
Time (days)
200
250
Plot Location: Middle
50
100 150
Time (days)
200
250
Plot Location: Toe
50
100 150
Time (days)
Subsoil-GYPSUM
200
GCL/Subsoil - FIBERGLASS
250
D-9
-------
PLOTJ
Bentomat - Granular Drainage-2:1 Slope
Moisture Readings vs. Time
Plot Location: Crest
50
100 150
Time (days)
200
250
100 -
90 -
80 -
70 -
60 "
50 -
40 -
30 *
20 -
10 -
0 -
•A A A.
50
Plot Location: Middle
100 150
Time (days)
200
250
50
Plot Location: Toe
100 150
Time (days)
Subsoil-GYPSUM
200
GCL/SubSOil - FIBERGLASS
250
D-10
-------
PLOTK
Claymax - Granular Drainage - 2:1 Slope
Moisture Readings vs. Time
Plot Location: Crest
50
100 150
Time (days)
200
250
50
100 150
Time (days)
200
250
Plot Location: Toe
100 150
Time (days)
Subsoil-GYPSUM
200
GCL/Subsoil - FIBERGLASS
250
D-ll
-------
PLOTL
Bentofix I - Granular Drainage - 2:1 Slope
Moisture Readings vs. Time
Plot Location: Crest
100 150
Time (days)
200
250
Plot Location: Middle
|T •*-•—•-»
100 150
Time (days)
200
250
Plot Location: Toe
100 150
Time (days)
Subsoil-GYPSUM
200
GCL/Subsoil - FIBERGLASS
250
D-12
-------
PLOTN
Bentofix II (NW up) - 2:1 Slope
Moisture Readings vs. Time
Plot Location: Crest
50
100 150
.Time (days)
200
250
Plot Location: Middle
50
100 150
Time (days)
200
250
100 :
Plot Location: Toe
50
Subsoil-GYPSUM
100 150
Time (days)
GCUSubsoil -
FIBERGLASS
200
w/in GCL -
FIBERGLASS
250
D-13
-------
PLOTP
Gundseal (bentonite up) • 2:1 Slope
Moisture Reading
|
i
Moisture Reading
Moisture Reading
Moisture Reading
i
100 •
90
80
70
60
50
40
30
20 !
10 * __
0
100
90
80
70
60
50
40
30
20
T —^
0
100 •
90 -
80 •
70 •
60 t
50 •
40 •
30 •
20 •
10 •
0-0-==:
0
100 -
90 -
80 •
70 «
60 j
50 I
40 :
30 ;
20 '
10 ' ,
0> R " r
0
— • Inner lett
panel
Moisture Readings vs. Time
Plot Location: Crest
^? P — ~ ~ — ' — ff
10 20 30 40 50
Time (days)
„
Plot Location: Upper Middle
10 20 30 40 50
Time (days)
Plot Location: Lower Middle
10 20 30 4.0 50
Time (days)
Plot Location: Toe
10 20 30 40 50
Time (days)
outer left • Inner right 0 outer right
panel panel panel
I
i
|
<
i
i
!
D-14
-------
Appendix E
Results of Interface Shear Tests Performed at Drexel University's Geosynthetic Research
Institute
-------
-------
GUNDSEAL (foentonite side) vs HOPE (textured) - DIRECT
SHEAR (12")
300
250
200
s 15°
oc
§ 100
W 50
0
0.0 0.2 0.4
350 Ibs (2.6 psi) Normal
Stress
I ..(.._..
0.6 0.8
DISPLACEMENT (IN)
1.2 1.4 1.6
a.
tn
JC
CO
234
Normal Stress (psi)
E-l
©
-------
BENTOMAT (slit film) vs HOPE (textured) - DIRECT SHEAR
(12")
1 OU •
140
120
100 •
80 -
60
40
20
01
•
0
•
•
M
'••
•
•
•
,, I
0 0.2 0.4
350 Ibs (2.6 psi) Normal
Stress
0.6 0.8
DISPLACEMENT (IN)
1.0
1.2
1.4
CO
a.
CO
CO
Q)
&
t_
(0
CO
4-
3-
2-
0
Q Peak
• Residual (@ 1.4" displ.)
2 3
Normal Stress (psi)
E-2
4
slid 1/12/95
-------
SHEARPRO vs TEXTURED GEOMEMBRANE
1.20 -
-. 1.00 •
S.
u 0.80
O
g 0.60
cc
52 0.40
0.20 ,
0.00
0.00 0.20
2.6 psi Normal Stress
0.40 0.60 0.80
DISPLACEMENT (In)
1.00 1.20
2.5
(0
^Q.
1
O
2.0-
y ~ 0.36502X RA2 = 1.000
T
2 3 4
Normal Stress (psi)
slid 12/10/95
E-3
-------
BENTOFIX (NPNWGT) vs HOPE (textured) - DIRECT SHEAR
(12")
300
^ 250 •
3
jJ-200
p
p 150
u§ 100
W 50
0
350 Ibs (2.6 psi) Normal
Stress
0.0 0.2
0.4
._ ,-4
0.6
0.8
DISPLACEMENT (IN)
1.0 1.2
1.4
to
o.
U)
V)
CO
Q)
0
234
Normal Stress (psi)
E-4
©
-------
200
180
co 160
,J 140
8-120
o 100
C 80
2 60
w 40
20
BENTORX-type II (Slit Film) vs HOPE (textured) - DIRECT
SHEAR (12")
0.0
350 Ibs (2.6 psi) Normal
Stress
0.2
0.4
. 0.6 0.8
DISPLACEMENT (IN)
1.0
1.2
1.4
w
Q.
V)
tO
2>
CO
to
-------
SHEARPRO VS SAND - DIRECT SHEAR (12")
250
sr200
150
100
50
2
•
I
y ^—.»T,-—.,..,—,,.. „ ,„,
0.0 0.2
0.4
350 Ibs (2.6 psi) Normal
Stress
0.6 0.8
DEFLECTION (IN)
1.0
1.2
1.4
0)
£
£
CD
sz
CO
y = 0.59696X RA2 = 1.000
2 3
Normal Stress (psi)
E-6
-------
SAND - DIRECT SHEAR (12")
200
350 Ibs (2.6 psi) Normal
Stress
0.0 0.2 0.4 0.6 0.8
DEFLECTION (IN)
1.0
1.2
1.4
o
u.
l_
(0
o>
CO
4-
3-
2-
1-
y = 0.51711x RA2 = 1.000
234
Normal Stress (psi)
E-7
-------
-------
Appendix F
Evaluation of Various Aspects of GCL Performance, Prepared by GeoSyntec Consultants
-------
-------
EVALUATION OF VARIOUS ASPECTS OF GCL PERFORMANCE
by
R. Bonaparte1, M.A. Othman1, N.R. Rad1, R.H. Swan1,
and D.L. Vander Linde1
INTRODUCTION
The purpose of this paper is to briefly present the results of various activities
that have recently been undertaken by the authors on the subject of geosynthetic
clay liner (GCL) testing and performance evaluation. The subjects that are
addressed are:
• field hydraulic performance of composite liners containing GCLs;
• drained shear strength of hydrated GCLs at high normal stress;
• interface shear strength between unhydrated GCLs and textured
geomembranes at high normal stress;
• hydration of GCLs adjacent to soil layers; and
• causes of failure of a landfill cover system containing a GCL.
FIELD PERFORMANCE OF COMPOSITE LINERS CONTAINING GCLs
Sources of Flow in LDS of Double-liner System
A double liner system consists of top and bottom liners with a leakage
detection system (LDS) between the two liners. If the double-liner system is used
in a landfill, it will also contain a leachate collection and removal system (LCRS)
above the top liner. As part of an ongoing research investigation for the United
States Environmental Protection Agency (USEPA), the authors have collected data
'GeoSyntec Consultants, 1100 Lake Hearn Drive, Atlanta, Georgia 30342
F-l
-------
on the rates of liquid flow into the sumps of LCRSs and LDSs for a wide variety
of double-lined waste management units located throughout the United States.
Comparison of the rates of flow into the LCRS and LDS of a unit can be used to
quantify the performance of the top liner (in terms of the ability to impede
advective transport of liquid through the liner). In essence, the LDS serves as a
large lysimeter (i.e., collection pan) below the top liner.
To make the evaluation, consideration must be given to the potential sources
of liquid in the LDS. Gross et al. [1990] described the potential sources of LDS
flow, which are (Figure 1): (i) leakage through the top liner; (ii) drainage of
water (mostly rainwater) that infiltrates the leakage detection layer during
construction but does not drain to the LDS sump until after start of facility
operation ("construction water"); (iii) water expelled from the LDS layer as a
result of compression under the weight of the waste ("compression water"); (iv)
water expelled from any clay component of the top liner as a result of clay
consolidation under the weight of the waste ("consolidation water"); and (v) for
a waste management unit with its base located below the water table, groundwater
infiltration through the bottom liner ("infiltration water").
Gross et al. [1990] and Bonaparte and Gross [1990] presented the following
five-step approach for evaluating the sources of LDS liquid at a specific waste
management unit.
• Identify the potential sources of flow for the unit based on double-liner
system design, climatic and hydrogeologic setting, and unit operating
history.
• Calculate flow rates from each potential source.
• Calculate the time frame for flow from each potential source.
• Evaluate the potential sources of flow by comparing measured- flow
rates to calculated flow rates at specific points hi time.
• Compare LCRS and LDS chemical constituent data to further establish
the likely source(s) of liquid.
Bonaparte and Gross [1990, 1993] used this five-step approach to evaluate
the sources of LDS flow for 93 waste management units. Under a contract to the
USEPA. Risk Reduction Research Laboratory, the authors are currently
performing this evaluation using new data from the facilities in the Bonaparte and
Gross studies, as well as data from a significant number of additional waste
management units not included in the original studies. Preliminary results for
waste management units with composite top liners containing GCLs are presented
below.
F-2
-------
GEOMEMBRANE
GROUND-WATER
TABLE
Q* TOTAL FLOW
Q=A+B+C+D
SOURCES:
A = TOP LINER LEAKAGE
B « CONSTRUCTION WATER AND COMPRESSION WATER
C « CONSOLIDATION WATER
D » WATER FROM GROUND-WATER INFILTRATION
Figure 1. Sources of flow from leak detection layers.
LCRS and LDS Flow Data
Flow rate data have been collected for 26 waste management units containing
composite top liners consisting of a geomembrane overlying a GCL. The 26
units are located at six different landfill sites. Descriptions of the components of
the liner systems used at these facilities are presented in Table 1 and flow rate
data for the LCRSs and LDSs in the units are presented in Table 2. Average
daily flow rates were calculated for both systems on a monthly basis by dividing
the total amount of liquid extracted from the system during the month by the
number of days in the month and the area of the waste management unit. Flow
rates are reported in units of liter/hectare/day (Iphd). The volume of flow used
in the calculation was typically obtained from the landfill operator, with flow
measurements most often measured using accumulating flow meters. The
reported flow volumes should be considered approximate.
F-3
-------
0)
c
o
I
u
1
CU
'S
*_.
j
n
u.
U
Fr
H
u
1
o
0
id
§
^
cs
L-
.3
£
u
ac
to
W3
c
^o
H
"«
L.
0>
m
s
C )
O
a
e~
u
S
Thickness
«
*n
S
§
1
r
d)
s
it
H
cd
i
H
U U
E V
o '^
I
5 f
.y E
4= "^
H
o *^
fe S
F %
E c
^
I
§8=18-
>. >, J <»• >, J
ra rt rj ^i rt M
u u S £ u 5
V) O tO «O VI O
.— « (Sj — * ~^ ^M .— I
So o
ir> o 10 10 10
_ _ „ ^ z z
1 1 13 § o o
-^2 sS _rr *•"•- LM C_H
o ° °
ro ro co co r*^ m
in 0 >n
-------
i
O O
O O
O
CJ
o o o> o
dodo
Sfl »-* —
o e o
I
*
II
f-£
— md — d-*dddddddd
°i °.
d d
£
(=(£
OvOOOOOOO
~ d o
d •«•' z:
..
£ " t-" £2 «n P-" 2 s en 2 2 »" 2 « — j;
< <£
— • m in ^» «- IN"
I is
\o oC 12
111
3 = 6
o jj E
5M U
_ s«
in ^ E
5 S
O CO 00
Is I
[5 U t j
S S! Si
S!
o O
U Z
50
5SS
£ C
F-5
-------
Figure 2 shows LCRS and LDS average daily flow rate data for a municipal
solid waste management unit, located in Pennsylvania, that was active for 56
months. Subsequently, a final cover system containing a geomernbrane was
placed over the entire unit. Flow data for the 56-month operational period and
a 25-month post closure period were obtained and analyzed. As Figure 2 shows,
flow rates in both systems were highest immediately after the start of waste
placement and thereafter decreased with time. During the first twelve months of
operation, the average rate of flow into the LCRS sump decreased from 12,700
to 180 Iphd. After that time, the LCRS flow rate stabilized and during the
following 44 months, the rate of flow into the LCRS sump varied between 10 and
170 Iphd. After final closure, the flow rate decreased even further, to between
10 and 80 Iphd.
As illustrated in Figure 2, waste management unit development can be
divided into three distinct periods. During the first period, herein referred to as
the "initial period of operation", LCRS flow rates may be relatively high. High
flows during this period are attributed to the occurrence of rainfall into a unit that
initially contains little waste. To the extent rainfall occurs during this period, it
will find its way rapidly into the LCRS. Obviously, the amount of LCRS flow
during this period is highly dependent on climate. A lag exists between the time
liquid first enters the LCRS and when it flows into the LCRS sump, The
magnitude of the lag is largely dependent on the hydraulic characteristics (i.e.,
the length and slope of the LCRS and the hydraulic conductivity of the LCRS
drainage material). Most available data indicate a decreasing LCRS flow rate
with time during the initial period of operation. During the second period,
referred to herein as the "active period of operation", the rate of flow into the
LCRS continues to decreases and eventually stabilizes. This occurs as the amount
of waste in the unit increases and as daily and intermediate layers of cover soil
are placed. This trend in flow rates is also dependent on the type of waste but
is likely representative of the trends observed at most new landfills, excluding
those that accept sludges or other high moisture content wastes. During the "post
closure period", the final cover system further reduces infiltration of rainwater
into the waste, resulting in a further reduction in LCRS .flow. Final covers
containing geomembranes can, if functioning properly, virtually eliminate
rainwater infiltration.
LDS flow rates for the waste management unit hi Figure 2 were highest (860
Iphd) at the beginning of operations and decreased hi the following few months,
becoming very low (i.e., less than 10 Iphd) within approximately 15 months after
the start of unit operation. The decrease in LDS flow with time is expected
because: (i) flow rates in the LCRS during this tune period decreased, and
therefore, the potential for leakage through the top liner also decreased; (ii) most
construction water initially present in the LDS flowed to the LDS sump in the
first few weeks to months of unit operation; and (iii) the volume of compression
and consolidation water for this waste management unit should be very small.
F-6
-------
c!
£
q 1000-
leu-
140-
120-
100-
80-
60-
40-
20-
0-
1
§
-
• •
lk
IP'
i *^* r"
L ^ i
• •.. , •• . ! , • , , , _, ! 1 , , ,J
eog> o> os o — — ew ™ 52 m 2 SiS !£
CD 00 CO' O> CO ' . CD O> ' O> O> O> O> CO CD D> p
^ j ? •• iv ^j^i^i^j^s^
DATE
Figure 2. LCRS and" LDS flow rates at a modern MSW landfill in
Pennsylvania.
F-7
-------
Table 2 summarizes LCRS and LDS flow data for the 26 waste management
units containing GCLs in their composite top liners. Average and peak flow rates
are reported for the three time periods described above. Table 2 shows that
between the initial and active periods of operation, LCRS flow rates decreased
one to two orders of magnitude and LDS flow rates decreased one to three orders
of magnitude. Reported peak LCRS flow rates were up to 5 times the average,
while peak LDS flow rates were up to 20 times the average. Table 3 presents the
mean values of average and peak flows for the database.
Table 3. Mean values of flow for the data in Table 2 (Note: m = mean value;
a = standard deviation; values are in liter/hectare/day).
LCRS
Initial Period of Operation
Active Period of Operation
Post-Closure Period
Number
of Units
25
18
4
Average Flow Rate
m
5,350
276
124
a
3,968
165
-
Peak Flow Rate
m
14,964
752
266
a
11,342
590
-
LDS
Initial Period of Operation
Active Period of Operation
Post-Closure Period
Number
of Units
26
19
4
Average Flow Rate
m
36.6
0.7
0.2
a
68.5
1.1
-
Peak Flow Rate
m
141.8
7.7
2.3
a
259.9
13.7
-
F-8
-------
Top Liner Hydraulic Efficiency
Table 4 summarizes calculated "apparent" efficiencies for the composite top
liners of the 26 waste management units presented in Table 2. Liner apparent
efficiency, AE, is calculated using the following equation:
AE (%) = (1 - LDS Flow Rate / LCRS Flow Rate) x 100 (Equation 1)
Table 4. "Apparent" efficiencies of composite liners containing GCLs(
Cells with Sand LDS
Cell No.
Al
A2
A3
A4
A5
A6
A7
A8
A9
A10
All
A12
A13
A14
A15
A16
Bl
,,- Cl
C2
Number
Range
Mean
Median
Initial Period
of Operation
(%)
100.00
99.90 •
98.97
96.01 .
97.23
98.58
99.37 .
99.02
99.91
100.00
100.00
100.00
100.00
100.00
100.00
100.00
94.57
97.94
91.84
19
91.84 - 100.00
98.60
99.90
Active Period
of Operation
(%)
100.00
99.33
98.71
98.75
97.50
100.00
. 99.20
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
99.29
17
97.5 - 100.00
99.58
100.00
Post-Closure
Period
(%)
100.00
100.00
99.55
100.00
4
99.55 - 100.00
99.89
100.00
Cells with GT/GN LDS
Cell No.
Dl
D2
El
Fl
F2
F3
Number
Range
Mean
Median
Initial Period
of Operation
(%)
99.98
100.00
99.95
100.00
99.95
99.90
6
99.90 - 100.00
99.96
99.97
Notes: "' Apparent Efficiency = (1 - LDS Flow / LCRS Flow) x 100
F-9
-------
This liner efficiency is referred to as "apparent" because, as described above,
flow into the LDS sump may be attributed to sources other than top liner leakage
(Figure 1). If the only source of flow into the LDS sump is top liner leakage,
then Equation 1 provides the "true" liner efficiency. Liner efficiency provides
a measure of the effectiveness of a particular liner in limiting or preventing
advective transport across the liner.
Table 4 presents calculated AE values for waste management units with sand
LDSs (Landfills A, B, and C). For these units, the apparent efficiency is lowest
during the initial period of operation (AEm = 98.6 percent; where AEm = mean
apparent efficiency) and increases significantly thereafter (AE,,, = 99.58 percent
during the active period of operation and AEm = 99.96 percent during the post
closure period). The lower AEm during the initial period of operation can be
attributed to LDS flow from construction water. For units A, B, and C,
calculated AE values during the active period of operation and the post-closure
period may provide a reasonably accurate indication of true liner efficiency for
the conditions at these units during the monitoring periods. It should be noted,
however, that the true efficiency of a liner is not constant but rather a function
of the hydraulic head in the LCRS and size of the area over which LCRS flow
is occurring (the area is larger at high flow rates compared to low flow rates).
Table 4 also presents calculated AE values for waste management units with
geonet LDSs (Landfills D, E, and F). The available data are limited to the initial
period of unit operation. As shown in Table 4, AE,,, for the six units with geonet
LDSs is 99.96 percent. This value is much higher than the AEm of liners of cells
with sand LDSs for the same facility operational period (i.e., 98.60 percent).
This higher efficiency can be attributed to the differences hi liquid storage
capacity and hydraulic transmissivity between sand and geonet drainage materials.
A granular drainage layer can store a much larger volume of construction water
and releases this water more slowly during the initial period of operation than
does a geonet drainage layer. This suggests that, during the initial period of
operation, the mam source of flow in a sand LDS underlying a composite top
liner containing a GCL is construction water.
Conclusions on Field Performance of Composite Liners Containing GClJs
From Table 2, LDS flows attributable to top liner leakage very from 0 to 50
Iphd, with most values being less than about 2 Iphd. These flow rates are very
low. The data shown hi Table 4 suggest that the true hydraulic efficiency of a
composite liner incorporating a GCL may be greater than 99.90 percent. A liner
with this efficiency, when appropriately used as part of an overall liner system,
can provide a very high degree of liquid containment capability.
F-10
-------
SHEAR STRENGTH OF HYDRATED GCLs
Overview
For-a recent project, the authors were concerned with the long-term drained
shear strength of hydrated GCLs at normal stresses in the range of 240 to 720
kPa. Drained shear strengths are applicable to long-term design and the range of
considered normal stresses is applicable to conditions in a liner system at the base
of a landfill, A testing program to evaluate the long-term drained shear strength
of GCLs was undertaken and this program is ongoing. To develop interim values
for preliminary design, the authors reviewed and analyzed available data from the
technical literature on the consolidated-drained (CD) shear strength of GCLs.
The findings of this review are presented below.
Required Deformation Rates to Achieve CD Conditions
To achieve consolidated drained (CD) test conditions, direct shear tests must
be carried out at a very slow rate of shear displacement. The required
displacement rate can be estimated using the well-known time-to-failure equation
specified in American Society of Testing and Materials (ASTM) standard test
method D 3080:
= 50
(Equation 2)
where: if = total elapsed time to failure(s); and t;0 = time required for the test
specimen to achieve 50 percent primary consolidation under the specified normal
stress, or increment(s) thereof. Using tf from consolidation tests and an estimated
failure displacement 6,, the required shear displacement rate, dr, can be calculated
using the equation:
d, = 6t /
(Equation 3)
Shan [1993] performed one-dimensional consolidation tests on the GCL products
Claymax*, Gundseal*, Bentomat*, and Bentofix*. He evaluated tf values for each
product. The results of his evaluation are provided in Figure 3. With reference
to this figure, at normal stresses in the range of 240 to 720 kPa, tf values are hi
the range of about 100 to 400 hours. If it is assumed that a displacement of 25
mm is needed to achieve peak shear stress conditions, a required shear
displacement rate of 0.05 to 0.25 rnm/s is calculated. Only test results conducted
at shear displacement rates that satisfy Equations 2 and 3 and the data from Shan
[1993] should be considered to represent CD conditions. Test results at faster
rates will yield lower shear strengths as a result of positive pore pressure
development during the shearing phase of the test.
F-ll
-------
600
.-. 500
LU
UJ
P
i
z
400
300
200
100
i"* BENTOMAT®
GUNDSEAL®
CLAYMAX®
BENTOFIX®
NORMAL STRESS (psi)
Figure 3. Relationship between time to failure of GCLs in direct shear tests
and normal stress (from Shan, 1993; Note: 1 psi = 6.9 kPa).
It is noted that direct shear tests on GCLs are often performed hi general
accordance with the standard test method ASTM D 5321 ("Determining the
Coefficient of Soil and Geosynthetic or Geosynthetic and Geosynthetic Friction by
the Direct Shear Method"). This method provides the following guidelines for
selecting shear displacement rates for tests involving soils:
"11.6 Apply the shear force using a constant rate of displacement that is
slow enough to dissipate soil pore pressures, as described in Method D 3080
(Note 9). If excess pore pressures are not anticipated, and in the absence
of a material specification, apply the shear force at a rate of 1 mm/min (0.04
in./min)."
F-12
-------
The foregoing requirement calls for performing direct shear tests involving soils
at a shear displacement rate in conformance with ASTM D 3080 if pore pressures
are anticipated. For the soil component of GCLs (i.e., sodium montmorillonite),
significant pore pressures will certainly be generated if the GCL is sheared at
rates faster than those satisfying Equations 2 and 3. Interestingly, however, most
test data available in the published literature were generated at the default shear
displacement rate of 0.017 mm/s. Data generated at the default shear
displacement rate are considered to reflect "undrained" or "partially-drained," and
not "fully-drained," conditions.
Review of Available Information for Unreinforced GCLs
For purposes of shear strength characterization, two different categories of
GCL can be considered: GCLs that do not contain internal reinforcement
(hereafter referred to as unreinforced GCLs) and those that do (hereafter referred
to as reinforced GCLs). Published information relevant to the CD shear strengths
of unreinforced GCLs is very limited. The available information is summarized
below.
• Daniel and Shan [1991] and Shan and Daniel [1991] reported CD direct
shear test results for the GCL product Claymax*. Tests were performed
using 60-mm diameter specimens and a shear deformation rate of 5 x
10"6 mm/s. Test results have been interpreted herein in terms of "peak
(p)" and "large-displacement (Id)" normalized shear strengths. Peak
displacements in these tests were 0.5 to 5 mm with the largest
displacement corresponding to the lowest normal stress; the reported
"Id" shear strengths correspond to shear displacements of approximately
6 to 9 mm. Results from the tests are as follows:
0.236
0.238
0.194
0.178
13.3°
13.4°
11.0°
10.1°
0.236
0.209
0.165
0.137
13.3°
11.8°
9.4°
7.8°
where: an = normal stress on the shear plane at failure (kPa); T =
shear stress on the shear plane at failure (kPa); and = secant friction
angle (dimensionless), calculated as the inverse tangent of r/an. It is
noted that should also be interpreted as a measure of normalized shear
strength and not as a "true" indication of internal friction. This data
interpretation is illustrated in Figures 4 and 5.
F-13
-------
-6^40 log (o „) + 25.877; f = 0.644
Claymax
Gundseal
95% standard error range
67% standard error range
S S g g g 8
NORMAL STRESS, on (kPa)
Figure 4. Log-linear regression analysis for peak CD conditions.
-8.867 log (o „) + 27.637; f = 0.902
Claymax
Gundseal
- 95% standard error range
67% standard error range
Figure 5.
S g § g 8
v-
NORMAL STRESS, on (kPa)
Log-linear regression analysis for large-displacement CD
conditions.
F-14
-------
Daniel and Shan [1991], Daniel et al. [1993], and Shan [1993] reported
direct shear CD test results for the GCL product Gundseal*. Tests were
performed using 60-mm diameter specimens and a shear deformation
rate of 5 x Itt6 mm/s. Test results have been interpreted herein in
terms of peak and large-displacement normalized shear strengths.
Typical peak displacements in these tests were 2 to 4 mm with the
largest displacement corresponding to the lowest normal stress; the
reported "Id" shear strengths correspond to shear displacements of
approximately 9 to 12 mm. Results from the tests are as follows:
J&u-
13.0°
12.8°
9.6°
9.3°
27
44
61
100
140
0.275
0.300
0.256
0.223
0.181
15.4°
16.7°
14.4°
12.6°
10.3°
0.231
0.227
0.169
0.164
The direct shear test results from Daniel and Shan are plotted in Figures 4
and 5 for "peak" and "large displacement" shearing conditions, respectively.
Regression equations were developed to describe the test results. It is interesting
to note the lesser amount of scatter in the results for the large-displacement
shearing conditions compared to the peak shearing conditions.
The test results in Figures 4 and 5 only cover the stress range between 24
and 144 kPa. Even at these relatively low normal stresses, GCL CD shear
* strengths exhibit significant normal stress dependency. A basis is needed for
extrapolating this stress dependency to higher normal stress. This basis was
derived from published information from the soil mechanics literature on the
shear strength of sodium montmorillonite. This information is summarized
below.
• Mesri and Olson [1970] and Olson [1974] reported the results of
constant rate-of-strain CD and consolidated-undrained (with pore
pressure measurement) triaxial compression tests on homipnic sodium
montmorillonite consolidated from a slurry (Figure 6); approximate
effective-stress normalized shear strengths and secant friction angles
derived from the tests are as follows:
gn (kPa)
72
170
340
530
0.21
0.14
0.10
0.07
12°
8°
6°
4°
F-15
-------
Ci
Figure 6. Effective-stress failure envelopes for calcium and sodium
montmorillonite from CD and CU triaxial tests (from Mesri and
Olson, 1970; Note: 1 psi = 6.9 kPa).
• Mitchell [1993] presented residual shear strength data for
montmorillonite from Kenney [1967] and Chattopadhyay [1972].
Inspection of the residual shear strength data shown in Figure 7 reveals
several significant points:
the residual friction angle exhibits significant stress dependency
over a wide range of normal stress; stated differently the residual
failure envelope is curved over a wide range of normal stress;
there may exist a normal stress above which the residual friction
angle is independent of normal stress; based on Figure 7, this
normal stress may be on the order of 480 kPa for sodium
montmorillonite; and
the residual friction angle of montmorillonite is dependent on the
dominant exchangeable cation and the soil pore chemistry; the
smallest measured residual friction angle given in Figure 7 is 3°
for homionic sodium montmorillonite in distilled water.
The GCL regression lines from Figures 4 and 5 are plotted along with the
Mesri and Olson [1970] data in Figure 8. Reasonable agreement is observed
between the Mesri and Olson data and the extrapolated regression lines for the
unreinforced GCL. Also shown on this figure are the residual shear strengths for
sodium montmorillonite developed by Kenney [1967] and Chattopadhyay [1972]
as reported by Mitchell [1993]. These latter results further support the
extrapolations presented in Figure 8.
F-16
-------
0.7
0.8
KEY
(1) DATA FROM KENNEY (1967)
(2) DATA FROM BISHOP elal. (1971)
(3) DATA FROM CHATTOPADHYAY (1972)
(2) ATTAPULGITE
40.1 gmAKer NaCI
(1) Na-HYDROUS MICA
(2)WRAYSBURYCLAY
(I)KAOLINITE
(3) MONTMORIUONITE
• 34 gm/Wer Nad
(2) WEALD CLAY
(1) Na-MONTMORILLONITE
0 gm NaCI
(onV1/3 (on'inpsi)
Figure 7. Residual effective-stress friction angles for clay minerals (from
Mitchell, 1993; Note: 1 psi = 6.9 kPa).
»_
1R «
16-
14 -
12-
10-
8-
6 —
^ _
2_
„
^
'""•-
h^
•^i,
••*••
"* •»
«»»^
•--v.
^^
-«i:
-~^
'"
*-.
^
•"*-
' in.
^*
•~-K
— .
-,
-«,
•~»
...
^
' —
1
___,
0
^..••^.^
^^-^^_
"^^^0^
*""•••
Pt«*(Figure4)
Laige Dtopteeement (Figure 5)
Moori and Obon [1970] data
Wontmorilonrte, 0 grammar MaCI (Figure 7)
MontmorllonJIe. 34 grmnVUer NaO (Figure 7)
•»
""*•-.
"^-^
^
"•"— -..^
' ^ ^
-^
~~-^.
•~-^,
1T.~;.'!
'~.
.
f\,. ,
^
l.«^w
^ *
->
•— ,
,
'
_.
8
^
NORMAL STRESS, on (kPa)
Figure 8. Comparison on montmorillonite shear strength data and GCL log-
linear regression lines.
F-17
-------
Review of Available Information for Reinforced GCLs
The authors were unable to find any information in the published technical
literature on the CD shear strengths of reinforced GCLs at high normal stress.
A few CD tests performed at low normal stress have been reported by Daniel and
Shan [1991] for the product Bentomat*. These results cannot be extrapolated to
higher normal stress, however, due to the current limited understanding of the
effect of reinforcing fibers on the shear displacement-shear resistance-normal
stress relationship for this type of material.
The authors have performed a limited number of consolidated-quick (CQ)
direct shear tests on reinforced GCLs at normal stresses in the range of interest.
Quick tests were performed at a displacement rate of 0.016 mm/s. While not
"truly undrained" due to the lack of boundary drainage control in the direct shear
test, the specimens in these tests will only undergo very limited pore pressure
dissipation during the shear phase of the test due to the high rate of shear
displacement. Due to these pore pressures, CQ tests at a given consolidation
stress will result hi lower GCL shear strengths than obtained from true CD tests
at the same normal stress. CQ tests may therefore be considered to provide a
lower bound of the CD shear strength of reinforced GCLs.
The results of the CQ direct shear tests on reinforced GCLs indicate
relatively high peak shear strengths followed, by a significant degree of shear
softening (i.e., post peak decrease in shearing resistance). A typical test result
is illustrated in Figure 9. Normalized peak and large displacement shear
strengths, and the ratio of the two (^) for a normal stress of 480 kPa are given
below:
Bentomat®
Bentofix®
Claymax® 500SP
29° 10°(i)
31° 16°(*)
13° 6°(l)
Xid/Ip
0.32
0.48
0.45
In the above table the downward arrow (I) indicates that the GCL shearing
resistance was decreasing at the end of the test (i.e., at a shear displacement of
40 to 50 mm). The $ values reported above are low, generally in the range of
0.3 to 0.5. In contrast, ^ values for the CD direct shear tests on unreinforced
GCLs were higher, typically in the range of 0.7 to 1.0. The 0ld values reported
above are somewhat larger than those obtained for the unreinforced GCLs.
However, as noted above, observation of the shear force-displacement plots for
the tests indicates that the shear stresses applied to the sample were decreasing
at the ends of the tests, which typically occurred at a displacement of 40 to 50
mm. This observation, coupled with observations of the tested samples, that the
GCL reinforcing fibers and stitching were still partially intact at the time the test
was terminated, suggests that residual CD and CQ shear strengths of reinforced
F-18
-------
GCLs may not be much larger than those of unreinforced GCLs. Clearly, testing
is required to establish the large-displacement, high normal stress behavior of
these materials, and to identify differences in product behavior based on
differences in montmorillonite properties and reinforcing characteristics.
o,, = 390 kPa
* On = 195kPa
0 on=100 kPa
LU
g
s
I I • I I I
20 25 30 35 40
10-
DISPLACEMENT (mm)
Figure 9. Results of CQ direct shear tests on reinforced Bentofix GCL.
Interim Design Values
Unreinforced GCLs: Based on the information presented in Figure 8, the
authors used the following interim guidelines for performing liner system stability
analyses for long-term drained conditions, for potential slip surfaces that involve
internal shearing of unreinforced GCLs. These guidelines further assume that the
GCL will hydrate through adsorption of water from an adjacent subgrade soil
layer.
• Slope stability analyses are performed using: (i) peak internal GCL
shear strengths and a minimum slope-stability factor of safety of 1.5;
and (ii) large-displacement internal GCL shear strengths and a minimum
slope-stability factor of safety of 1.15.
F-19
-------
Using the regression equation presented in Figure 4, peak normalized
shear strengths are:
gn (kPa)
96
240
480
720
0.214
0.157
0.114
0.106
12.1°
8.9°
6.5°
6.1°
Using the regression equation presented in Figure 5, large-displacement
normalized shear strengths are:
10.1°
6.6°
4.0°
3.0°
96
240
480
720
0.178
0.116
0.070
0.052
For the large displacement strengths, a minimum friction angle cutoff of 3°
was assumed based on the test results reported by Mitchell [1993], presented in
Figure 7.
The normalized shear strengths given above are relatively low, and their use
may be viewed by some as overconservative. This view should be tempered with
the realization that the large-displacement GCL shear strengths reported in the
technical literature do not represent true residual minimums (due to the limited
displacement of the direct shear apparatus) and no allowance has been made for
the possible effects of drained creep of the GCL under working stress conditions.
Furthermore, the available CD direct shear test results for unreinforced GCLs
correlate well with the triaxial compression test results for sodium
montmorillonite from Mesri and Olson [1970] and Olson [1974] (Figure 6).
Finally, it is noted that the foregoing approach, which utilizes a smaller slope
stability factor of safety with the large displacement shear strengths than die
factor of safety used with the peak shear strengths, is similar to the approaches
advocated by Byrne [1994] and Stark and Poeppel [1994].
Reinforced GCLs: Recognizing the lack of data on the CD strength of
reinforced GCLs at high normal stress, the complex behavior and high degree of
shear-softening exhibited by these products, the authors utilized the same factors
of safety and GCL long-term shear strengths for reinforced GCLs as for
unreinforced GCLs. It is recognized that this assumption is conservative.
However, given the limitations with respect to the available reinforced GCL test
data (e.g., the technical literature does not contain any "true" CD direct shear test
F-20
-------
results for reinforced GCLs at high normal stress) and the other factors discussed
above, the authors believe the assumption was prudent.
SHEAR STRENGTH OF GCL-GEOMEMBRANE INTERFACES
Direct Shear Testing Program
For a project located in the desert of southeastern California, the authors
performed 14 interface direct shear tests on unhydrated GCL-textured HDPE
geomembrane interfaces. The tests were performed in a 300 mm x 300 mm shear
box following procedures in general accordance with ASTM D 5321. Three
different GCLs were tested. The geomembrane used in the tests was from a
single roll of material and samples were selected based on visual observation of
a consistent degree of texturing. The tests were carried out in a manner that
allowed shearing either at the GCL interface or internally within the GCL
bentonite layer. Tests were carried out at normal stresses ranging between
approximately 350 and 1,920 kPa. Sliding in the tests consistently occurred at
the interface and not within the GCL. Thus, the test results correlate to interface
failures and at the same time provide conservative lower bound unhydrated shear
strengths for the tested GCLs under the project testing conditions.
Typical test results are presented hi Figure 10 and summarized in Table 5.
The tests correspond to two shearing rates, namely 0.016 mm/s and 0.0007
mm/s. Interface friction angles obtained from the tests at the slower shearing rate
are 1° to 2° lower than interface friction angles obtained from tests at the higher
shearing rate. The test results also reveal an interface shear strength stress-
dependency with secant interface friction angles 5° to 10° lower at 1,920 kPa
than at 350 kPa. The interfaces exhibited only minor amounts of shear softening
(typically less than 1 to 2°) at test displacements of up to about 50 mm.
Comment on Results
The foregoing interface direct shear test results illustrate the ranges of shear
strengths obtained and several of the factors that affect this strength including
normal stress, displacement rate, and magnitude of displacement.
The authors note that they have observed relatively wide variances in the
degree of texturing of geomembranes, even from a given manufacturer. The
degree of texturing significantly influences the interface shear strength. Thus, the
strength values reported above should not be considered appropriate for design.
Interface shear strengths for design should be established on a project-specific
basis and construction-phase quality control testing should be used to establish
that materials delivered to the construction site can achieve the interface strengths
established during design.
F-21
-------
Figure 10.
5 10 15 20 25 30 35 40 45 50 55 60
DISPLACEMENT (mm)
Results of direct shear tests on unhydrated Bentofix GCL-textured
HDPE geomembrane interface.
HYDRATION OF GCLs ADJACENT TO SOIL LAYERS
Overview of Testing Program
The authors conducted an extensive laboratory testing program to evaluate
the potential for hydration of GCLs placed against a compacted subgrade soil
layer. Hydration tests were performed on three different GCL products to
evaluate the effects of: (i) test duration (i.e., hydration time); (ii) soil initial water
content; (iii) thickness of soil layer; and (iv) overburden pressure. Three
commercially-available GCL products, namely, Claymax®, Bentomat®, and
Bentofix® were used in the testing program. The soil used in the testing program
was obtained from the USEPA GCL Field Test Site at the ELDA-RDF facility
in Cincinnati, Ohio. This material is classified as low plasticity clay (CL) based
on the Unified Soil Classification System (USCS). Tests were performed on two
different soil samples and consistent results were obtained between samples. The
results reported herein were obtained from tests on a sample with 99 percent of
the soil passing the U.S. No. 200 standard sieve and 33 percent smaller than 2
/*m (clay fraction). The liquid limit of the soil is 41 and the plasticity index is
19. The soil has an optimum moisture content (OMC) of 20 percent and a
maximum dry unit weight of 16.7 kN/m3 based on the standard Proctor
compaction method (ASTM D 698).
F-22
-------
05
1
i
2
2
£
1
too
w
OH
Q
ffi
O
00
CA
U
•S
I
S
W")
(U
1
S%
1 |j
5|S
f° S
c3
o
1
05
If
ll
81
1
Q
M
<8
T3 *•
l~
^
|
S
S
VO
•••«
0
0
o
•3
[nonwoven si
Bentomat GCL
S
VO
O
o
S
•8
[nonwoven si
Bentomat GCL
cs
VO
O
O
1
•o
us
O
i
Bentomat GCL
8
g
g
0
S
o\
•g
[nonwoven si
1 Bentomat GCL
2
vg
o
o
1
•S
CA
1
*-•
J5
^
g
o
1
•o
§
1 Bentomat GCL
CO
ts
VO
0
o
o
"aT
*o
'55
i
^5,
Bentofix GCL i
8
VO
O
O
VO
3?
lonwoven sid
&
Bentofix GCL i
00
5Q
VO
O
o
1
•«?
lonwoven sid
bM
Bentofix GCL i
Ov
ts
VO
O
o
1
"to*
*o
c
Bentofix GCL I
o
Tt
VO
0
o
o
/-v
IS
*>
"3
S
1
^*"
8
o
"•* -
S
VO
O
O
§
u
!2
S3
"3
00
Sj
1
s!i'
I
W-H
55
VO
O
O
1
,—v
'S3
S
1
ob
^2*
Gundseal GCL
"*
z
VO
0
o
I
^— ^
•a
'S3
(8
"3
bentonite gra
^>
73
1
•*
^
I'
1
_)
1
£
i
i
&o
1
3
00
_c
r*
«
1
0
^
"tt
C3
rS
ii e
u E
1 S
« 2
•5 s
S *o
CO 113
J &
8 1
S s
2 u: c
•v} U
£&£
c ~" «*
3 « S
a>'S a
S o «
II!
15*
ld.s
«g® a
« S S
c
2'? I1
S sl
.S'Sl
P s£
-^N X-S
O- s^
is
F-23
-------
Testing Apparatus and Procedure
Figure 11 shows the apparatus specially designed to conduct the GCL
hydration tests. The apparatus consists of a polypropylene mold 75 mm in
diameter and 150 mm hi height. A geomembrane/GCL/soil composite specimen
is placed in the mold and covered with two layers of a thin vapor barrier. A
loading platen is placed on the specimen for application of overburden pressure.
To process the soil, it was first passed through a U.S. No. 4 standard sieve.
The soil was then moisture conditioned to achieve the desired moisture content.
The moist soil was placed in the mold in a loose condition and statically
compressed to 50-mm thick lifts. The soil was compacted to a dry unit weight
equal to approximately 90 percent of the maximum dry unit weight based on the
standard Proctor method (ASTM D 698). Two soil lifts were used giving a total
thickness of 100 mm. The GCL and geomembrane specimens were carefully
trimmed from the same sheets. The initial moisture content of the GCL was
measured by taking a small sample from the same GCL sheet and measuring its
weight before and after oven drying. The initial moisture content of the GCLs
varied between 15 and 20 percent.
DOUBLE LAYER
MOISTURE
BARRIER -\
GEOSYNTHETIC
CLAY LINER
COMPACTED
SOIL
- LOADING WEIGHTS
(WHEN APPLICABLE)
LOAD PLATEN
- 40-MIL TEXTURED
HOPE GEOMEMBRANE
3" X 6" LENGTH
PLASTIC CYLINDER
Figure 11. Simplified diagram of GCL hydration test set-up.
F-24
-------
The GCL and geomembrane were placed on the soil and covered with the
vapor barrier. The side of the GCL placed against the soil was woven in the case
of Claymax® and nonwoven for Bentomat® and Bentofix®. Overburden pressure
of 10 kPa was applied on the composite specimen utilizing standard weights
which were placed on the loading platen. The entire apparatus was then placed
in a temperature and humidity controlled room for the desired hydration time
period. At the end of the hydration period, the test specimen was removed and
the water content of the GCL and soil were measured. The final moisture content
of the GCL was measured by weighing the entire GCL specimen before and after.
oven drying. The final moisture content of the soil was measured as the average
water content of three samples obtained from the top, middle, and bottom of the
soil specimen.
Testing Conditions and Results
As previously described, test conditions were varied to evaluate the effects
of several factors on the hydration of GCLs. To evaluate the effect of test
duration, tests were performed where the GCL was in contact with the soil for
.5, 25, and 75 days. Soil specimens were compacted to initial moisture contents
equal to OMC, 4 percentage points dry of OMC, and 4 percentage points wet of
OMC to evaluate the effect of soil initial moisture content on GCL hydration.
Figures 12, 13, and 14 present the results of the hydration tests for the GCL
products Claymax®, Bentomat®, and Bentofix®, respectively. These figures show
that the moisture content of all three GCLs increased significantly as a result of
contact with compacted subgrade soil. The increase in GCL water content was
significant after only five days of hydration. With increasing time, GCL water
content continued to increase at a decreasing rate. For most tests, GCL water
content reached a maximum value after about 25 days of soil contact and for
some of the tests water content continued to increase even after 75 days of
hydration. It is interesting to note that all three GCL products showed relatively
similar behavior. Increases in water content were comparable for the three GGL
products despite differences in GCL fabric (i.e., woven vs. nonwoven) and types
of bentonite clay used to manufacture the GCLs.
Figures 12, 13, and 14 illustrate the influence of soil subgrade initial
moisture content on the hydration of GCLs. From these figures, it is evident that
the moisture content of the GCL for any particular hydration time increases as the
initial moisture content of the soil increases. These figures also show that a small
increase in soil initial moisture content can have a significant impact on GCL
moisture content. For example, after 75 days of hydration, the moisture content
of Claymax® was approximately 16 percent higher when the initial moisture
content of the soil was equal to OMC than when it was 4 percentage points drier
than OMC. This behavior is expected because more water is available in the soil
for the GCL to hydrate. '
F-25
-------
100-r
80 - -
60-
! f
-44
..,
OMC + 4%
OMC
OMC-
Claymax (woven)
4%
40
—r-
60
80
100
HYDRATION TIME (DAYS)
Figure 12. Increase in GCL moisture content due to contact with compacted
subgrade soil: Claymax® with woven geotextile against soil.
fe
o
DC
1
100
80-
Bentomat (nonwoven)
40
60
80
100
Figure 13.
HYDRATION TIME (DAYS)
Increase in GCL moisture content due to contact with compacted
subgrade soil: Bentomat® with nonwoven geotextile against soil.
F-26
-------
LL
O
O
EC
1
100
80-
20
100
HYDRATION TIME (DAYS)
Figure 14.
Increase in GCL moisture content due to contact with compacted
subgrade soil: Bentofix® with nonwoven geotextile against soil.
The examination of the curves shown in Figures 12, 13, and 14 shows that
the time required for the GCL to reach its final moisture content is less in the
case of a dry soil than in the case of a wet soil. At the lowest soil'initial
moisture content tested, GCL moisture content ceased to increase after about 5
to 25 days. At the highest initial moisture content tested, the Bentomat® and
Bentofix® GCLs continued to increase in moisture content after 75 days of
hydration.
To evaluate the effect of soil layer thickness, specimens were prepared using
50, 100, 150, and 200 mm of soil thickness. Soil initial moisture content was 20
percent and dry unit weight was 14.9 kN/m3 for all specimens. Figure 15 shows
the results of hydration tests for the Bentofix® GCL after 25 days of hydration.
The GCL moisture content increased with the increase of the soil layer thickness.
However, it appears that only a small change in moisture content increase occurs
for thicknesses greater than 100 mm.
The effect of overburden pressure on GCL hydration is illustrated in Figure
16 for the Bentofix® GCL. As shown in this figure, overburden pressure in the
range of 5 to 390 kPa did not significantly affect the rate of GCL hydration
during the 25-day test duration.
F-27
-------
60-
g
ft
O
DC
40-
20
Figure 15.
i
i Bentofix (nonwoven)
—r~
50
T
—r~
200
100 150
SOIL SPECIMEN LENGTH (mm)
Influence of subgrade soil layer thickness on GCL moisture
content.
250
54-i
.
,m.
S> 52-
d
o
U. 50-
0
ULJ 48-
O
s <•-
LU
§
5 44-
»«.»»..»»
_
• ..•liii.i.im-T
! — !
-rr
.........
,...«.
i i Mi
*
...
M
_._i_.
!
i_4-
i
i
: : : :
i !!
...
! I
! !
.
i-4-
4
f.
T
4
i
j
i
!
1
i
!
JL
i
.
i 1
4
—
i i i ii j
"T !
i i s i :
i
!
...
""
! i
1 1
! i i i !
• 1 !
*•!...». . -.—......
..L
\
!
11
|
1"
i i
i
j. J. „
!
—
4
—
i
!_.«.
!
I
«..L...
i
i i i
i
i
.
!
j
„
••
"t"
i
t
_.
U-L.
Bentofix (nonwoven)
! i i
1
i ! Ii
Figure 16.
8
OVERBURDEN PRESSURE (kPa)
Influence of overburden pressure on the increase in GCL moisture
content.
F-28
-------
Summary
From the testing program results described above, the following can be
concluded:
• GCLs will hydrate when placed in contact with subgrade soils
compacted within the range of moisture contents typically found in
earthwork construction specifications; this conclusion is consistent with
data provided by Daniel et al. [1993]; even for the driest soil
(compacted 4 percentage points dry of OMC), GCL moisture contents
consistently increased from an initial value hi the range of 15 to 20
percent up to about 40 percent within a 100-day period; it should thus
be anticipated that GCLs placed even against relatively dry compacted
subgrades will undergo substantial hydration;
• given that Daniel et al. [1993] have shown that long-term GCL shear
strengths are inseasitive to water content for water contents above about
50 percent, stability analyses involving GCLs placed in contact with
compacted subgrade soils should be based on hydrated GCL shear
strengths;
• significant increases in GCL moisture contents may occur within a few
days of GCL contact with a moist soil; the rate of GCL hydration is
initially highest and then decreases with increasing tune;
• within the range of conditions tested a higher soil moisture content
results in a higher GCL moisture content;
• larger soil layer thickness results in a larger increase in GCL moisture
content,, however, for soil layer thicknesses greater than 100 mm only
. . increasing soil layer
insignificant increases
thickness;
were observed with
overburden pressure within the range tested (i.e., 5 to 390 kPa) did not
influence the hydration process; and
differences between GCL products tested (i.e., type of bentonite clay
and fabric) did not seem to significantly affect the test results.
F-29
-------
FAILURE OF LANDFILL COVER SYSTEM CONTAINING A GCL
Description of Cover System
The authors recently investigated the failure of a cover system for a
municipal solid waste landfill near Atlanta, Georgia. The failure is described in
more detail by Vander Linde et al. [1995]. The cover system was constructed in
the fall of 1994 on 3H:1V (horizontal:vertical) side slopes to a maximum height
above surrounding ground of approximately 18 m. The cover system consisted
of, from top to bottom:
• 300-mm thick layer of final cover soil which is classified as silty sand
containing approximately 40 percent fines based on ASTM D 2487, and
which has a hydraulic conductivity in the range of 1O4 to 10"3 cm/s;
• stitch-bonded reinforced GCL; and
• 150- to 300-mm thick layer of intermediate cover soil which served as
a foundation for the overlying final cover components.
Failure of System
During the winter of 1995, the cover system experienced several episodes
of downslope movement. The first major episode occurred approximately one
month after the completion of construction; the movement occurred after a three-
day period in which 58 mm of rain fell at the site. The next major episode
occurred six weeks later, after two days of inclement weather generated about 41
mm of rainfall at the site. Total downslope movements exceeded 1 m at some
locations. The observed failure mechanism was sliding of the final cover soil on
top of the GCL.
Analysis of Failure
The episodes of downslope movement both followed periods of extended
rainfall at the site. A slope stability back-analysis of the cover system was
performed which accounted for the influence of rainfall-induced seepage forces
on cover system factor of safety against downslope sliding. The back-analysis
involved two steps:
• estimating seepage forces within the cover soil using several different
calculation methods and parameter values; and
• calculating the resulting slope stability factors of safety for the range of
estimated seepage forces.
F-30
-------
The evaluation of seepage forces involved calculating the water build-up
(i.e., hydraulic head) within the final cover soil on top of the GCL. Head was
calculated using a methodology developed by Giroud and Houlihan [1995] and
checked using the United States Environmental Protection Agency (USEPA)
Hydrologic Evaluation of Landfill Performance (HELP) computer program
Version 3.03 fUSEPA, 1994a, 1994b]. The values of head calculated using these
approaches ranged from 150 mm to the full thickness of the cover soil layer, 300
mm.
Calculations to obtain slope stability factors of safety were performed using
the equations presented by Giroud et al. [1995a, 1995b]. An important input to
the equations is the shear strength of the interface between the cover soil and
GCL. Tests to evaluate the shear strength of this interface had not been carried
out as part of the original design. For the back-analysis of the failure, a range
of friction angles (20° to 26°) was considered for the cover soil-GCL interface;
this range likely brackets the actual interface strength and includes the value of
24° originally assumed by lie design engineer. Calculations were performed and
the following results were obtained:
Interface Friction
Angle (degrees')
20°
.24°
'26°
Factor of Safety (FS) vs. Hydraulic Head
0 mm 100 mm 200 mm
1.09
1,35
1.47
0.84
1.04
1.13
0.60
0.73
0.80
11 • " '
These calculation results demonstrate the significant impact of seepage forces
on the stability of the final cover soil. Even with the largest assumed interface
strength, only 140 mm of head buildup is required to decrease the slope stability
factor of safety to less than 1.0. Interface shear strength tests performed after the
completion of the back analyses resulted in peak and large-displacement secant
friction angles for the GCL-cover soil interface, at the applicable normal stress
of 23° and 21°, respectively.
Summary
The primary factor contributing to the observed final cover soil movements
was the build-up of seepage forces in the final cover soil during periods of heavy
rain. Seepage forces were riot accounted for in the design. If seepage forces had
been accounted for, the potential for instability likely would have been identified
during preparation of the design. The development of seepage forces in cover
soils is typically minimized by the inclusion of a drainage layer above the low-
permeability barrier component of the cover (in this case, the GCL). A
secondary factor contributing to the movements was a final cover soil-GCL
interface shear strength lower than assumed in the design. An interface friction
F-31
-------
angle of 24° was assumed by the design engineer, based on information provided
by the GCL manufacturer. The actual project-specific interface shear strength
was closer to 21°. This result highlights the fact that actual interface strengths
can only be assessed by project-specific testing; such testing was not performed
for the project.
ACKNOWLEDGEMENTS
The authors would like to thank Mr. Robert R. Landreth and Mr. David A.
Carson of the U.S. Environmental Protection Agency, Risk Reduction Research
Laboratory for their support hi the evaluation of double-liner system performance.
REFERENCES
Bonaparte, R. and Gross, B.A., "Field Behavior of Double-Liner Systems"
Proceedings of the Symposium on Waste Containment Systems, ASCE
Geotechnical Special Publication No. 26, San Francisco, CA, 1990, pp. 52-83.
Bonaparte, R. and Gross, B.A., "LDCRS Flow from Double-Lined Landfills and
Surface Impoundments", EPA/600/SR-93/070, U.S. Environmental Protection
Agency, Risk Reduction Research Laboratory, Cincinnati, OH, 1993, 65 p.
Byrne, J., "Design Issues with Strain-Softening Interfaces in Landfill Liners",
Proceedings, Waste Tech '94, National Waste Management Association,
Charleston, SC, Jan 1994, 26 p.
Chattopadhyay, P.K., "Residual Shear Strength of Pure Clay Minerals", Ph.D.
Dissertation, University of Alberta, Edmonton, Canada, 1972.
Daniel, D.E. and Shan, H-Y., "Results of Direct Shear Tests on Hydrated
Bentonitic Blankets", University of Texas, Geotechnical Engineering Center,
1991, 13 p.
Daniel, D.E., Shan, H-Y., and Anderson, J.D., "Effects of Partial Wetting on
the Performance of the Bentonite Component of a Geosynthetic Clay Liner,.
Proceedings, Geosynthetics '95 Conference, Vol. 3, Vancouver, Feb 1993, pp.
1483-1496.
Giroud, J.P. and Houlihan, M.F., "Design of Leachate Collection Layers",
Proceedings of the Fifth International Landfill Symposium, Vol. 2, Sardinia, Oct
1995, pp. 613-640.
F-32
-------
Giroud, J.P., Williams, N.D., and Pelte, T., "Stability of Geosynthetic-Soil
Layered Systems on Slopes", Geosynthetics International, Vol. 2, No. 6, 1995a,
pp. 1115-1148.
Giroud, J.P., Bachus, R.C., and Bonaparte, R., "Influence of Water Flow on the
Stability of Geosynthetic-Soil Layered Systems on Slopes", Geosynthetics
International, Vol. 2, No. 6, 1995b, pp. 1149-1180.
Gross, B.A., Bonaparte, R., and Giroud, J.P., "Evaluation of Flow from Landfill
Leakage Detection Layers", Proceedings of the Fourth International Conference
on Geotextiles, Vol. 2, The Hague, Jun 1990, pp. 481-486.
Kenney, T.C., "The Influence of Mineralogic Composition on the Residual Shear
Strength of Natural Soils", Proceedings of the Oslo Geotechnical Conference on
the Shear Strength Properties of Natural Soils and Rocks, Vol. 1, 1967, pp. 123-
129.
Mesri, G. and Olson, R.E., "Shear Strength of Montmorillonite", Geotechnique,
Vol. 20, No. 3, 1970, pp. 261-270.
Mitchell, J.K., "Fundamentals of Soil Behavior", 2nd Edition, John Wiley &
Sons, Inc., New York, NY, 1993, 437 p.
Olson, R.E., "Shearing Strengths of Kaolinite, Illite, and Montmorillonite",
Journal of the Geotechnical Engineering Division, ASCE, Vol. 100, No. GT11,
1974, pp. 1215-1229.
Shan, H-Y, "Stability of Final Covers Placed on Slopes Containing Geosynthetic
Clay Liners", Ph.D. Dissertation, University of Texas, Austin, TX, 1993, 296 p.
Shan, H-Y. and Daniel, D.E., "Results of Laboratory Tests on a
Geotextile/Bentonite Liner Material", Vol. 2, Proceedings, Geosynthetics '91
Conference, Atlanta, GA, Feb 1991, pp. 517-535.
Stark, T.D. and Poeppel, A.R., "Landfill Liner Interface Strengths from
Torsional-Ring-Shear Tests", Journal of Geotechnical Engineering, ASCE, Vol.
120, No. 3, 1994, pp. 597-615.
USEPA, "The Hydrologic Evaluation of Landfill Performance (HELP) Model,
User's Guide for Version 3", EPA/600/R-94/168a, U.S. Environmental Protection
Agency, Washington D.C., December 1994a.
USEPA, "The Hydrologic Evaluation of Landfill Performance (HELP) Model,
Engineering Documentation for Version 3", EPA/600/R-94/168b, U.S.
Environmental Protection Agency, Washington D.C., December 1994b.
F-33
-------
Vander Linde, D.L., Luettich, S.M., and Bonaparte, R., "Lessons Learned for
Failures of a Landfill Cover System", Geosynthetics: Lessons Learned from
Failures, J.P. Giroud and K.L. Soderman, Eds., International Geosynthetics
Society, 1995, in press.
F-34
-------
Appendix G
Summary of Bentomat Direct Shear Data, Prepared by CETCO
-------
-------
H
^^H
Q
H
H
*
^•^
£
vX)
H
U
C<
C
H
<
O
C~ '
^Z,
§
r
O
^
^Tj
^•^
^^j
S
£
V]
C
E
E
o
U
O
VI
C.
||
C. f
Q. 0
< U
C
*£Z
^ |f
"« 2-
^S
W3 OX
U B
cs: <
o ,_,
'•S M
_U u
IE s
^ Ji
Cu <
"S
si
j: c
C/3 O
11
5 "K
g 1
CQ .E
/_N
V]
7S S
E 2
o i
Z to
*>4
•o
•£*
VI
V
H*
Interface
*;
O v
Q. -^
u re
K Q
la
n
,
o m
— OO
•g
_c
OJ
"o
•o
o
in oo
m CN
CN fN
o o
o o
T3
u
2
">> b
I Q
fi r^
tN fN
i i
NW/Sand
NW/Sand
0
OS
o
fl
n
r~ o
5 S
— CN
TT fl
fN rN
O O
O 0
1
2
?>b
I Q
r~> ro
CN fN
i i
5 5
7. "Z.
o
"8
_n
o>
o
T3
O
00
CS
O
2^
Q
0
r-
cs
m
>n
a.
Q
?<
fM
r™
rzi
NW/40-m
o
5!
^"
ON
0
CO
CO
o o
5 5
r- •q-
fl fN
CN CN
0 0
b b
Q Q
o o
r-- r—
fN fN
•n vn
t i
in in
d. m
Q eu
I Q
X ^j
f^ X
*""
m t"*
I'f
85 S
OS
"8
_c
u
u
•o
B
CN
fN
O
O
b
Q
vo
i
f*^
i
fN
—.
*3
C/l
>s.
NW/Sand;
o
VO
o
_!.
"
J
2
1
o
«
-Si
.£
•a
rt
"O
rr"
oo m f>
CN — —
•2
c
1* »
4j
U
T3
O
fN r- os
o o o
o o o
T3 -0
Q> O
2 2
b'SL'5.
a re i
o
CN O O
O ' '
— in in
iii
1 ' '
*— '
III
.
US
_
OS
OO
•™
1
o
t
oi
a
o o
•g
c
2 "
u
•a
o
O OS
fN —
0 0
•a -o
w o
2 2
">."§,
i D:
o o
CN fN
•n in
in in
m f>
Oi tlj
Q a.
I Q
u SC
s -'
t- jo
= f-
5 o
' _
OS
CO
tN
i
•n
o
CO
CO
OS
vo
T3
_C
%
•a
o
VO
CN
rn
o
0
o
o
•a
2
"5,
I
OS
1
—
1
Os
1
VO
Internal
_
OS
CN
OO
o
<
=>
•n TT o
in vo —
O vo fN
CN — in
m
CN r~- r^
CM — m
•n in in
O O 0
o o o
•O TJ T3
8> U U
222
•a 13 -a
III
00 OO OO
OO OO OO
• ' *~" ^—
10 ir» vo
rs ^ O
i is
ill
Os
OS
O
oV
o
J
2
c
o
E
"^
j: -
•o
»
E
J
— r-
"S
c
u
u
•o
o
CN in
fN —
O 0
o o
T3 T3
2 2
x'l'
_ —
00 OO
1 1
CN fN
SI
^
t i
X g
£ t/1
11
o o
VO vo
fN
VO
er>
•n
O
5
o
fN
m
VO
vn
CN
CN
O
•o
. cd
•o
I
i
i
>n
rS
ID
O-
D
^
t •
X
{P
'§
^
CN
Os
fN
1
^"
5
CO
a:
ec
E
O
c
m
o m
o m o
— — fN
"S
c
V
o
•o
0
TJ- o r-
tN fN —
S^T "^
0 0
o o o
t> 0
2 2
T3 ID >,
n: a: c
n f> rn
fN CN CN
i i i
— — •—
O.
1
_ 8
'S OX)
, O "*~
1-1 OJD
^ — n
W/Saturati
W/Dry soi
NW/Drain
n
Os
vo
cA
O
5
G-l
-------
^r
.s
•*-»
c
o
U
^^L
^^
5j<
ft
H
t-
(•^
<
a
CO
£•4
Fj
fVI
W
2
MATDI
O
pK
w
fa
o
1:
0
E
E
o
U
o
WS
&
! ^
&"o
< U
c
o
s
£ 'M
"5 S
3 77
t3 H
8 =
OS <
c
.2 <—•
i£ f.
cs "^
£ <
u
£§ «-»
1* *"""
S E
CJ ~-»
CO O
11
B|
ca S
o
S.
Normal
Stresses
*TJ
V
I
V
1
u
£
O «>
S"S
« P
To
J3
"ra
3
"O
1
IM
"«
"8
E
15
ON
O (N
vo
OO
t- CS
oo
00 CNl
O 0
o o
•O T3
O CJ
2 2
T3 13
o o
^™ — *
t 1
**r ^>
J CN
i i
O O
o*
1
s>
ra
.5 —
£ £
"w
O.
O
oo
ra
S
*-*
X
1
ON
oo
0
o
Q
o
oo
1
s
1
o
TJ-
tl3
CL
Q
I
o
oo
&
r~-
9
o
P
— 00
00 TJ-
1
'1 ,
s
^3
•o
o
in
- S
0 0
o o
•o
2
i i
• i
U U
0. 0.
Vt V)
1 1
6 6
^ ^
?
0
ti
o
p
§o
tr» ^
c^i r5 —~ »o '—•' o
-O T3
Gi &>
c c
S S
o o
T3 T3
O O
"Z. in r- Z
in m
r~ ON tr — ' CN
00 0 O O O
O O 0 O O O
•a 13 -a
U O "?, b £>
D I Z I Q Q
o
tn ^—
i M" (N
_, i 0 tj- O O
M fv! r- — oo oo
1 ( i i it
"J" _ , £> NO
NM ^M !^ kM 2 !^
f*^ f^l "5f ^f
1 III
OO •—
1
u
CO
O
G-2
-------
•o
s
"•C
c
o
U
^^
H
Q
S
U
H
A
&
ffl
H
ca
C£
S
H
•<3
5
O
H
W
CQ
0
£H
04
^^
S
§
CO
-52
c
v
E
E
o
U
c-
VI
O,
1.1
u »
69 Q)
e. js
CL o
< U
o
T3
£ Q
"w r^.
3 ^
8 I1
OS <
B
.5 *-*
W U
£ ^
C8 S^
a! <
4J
01 ^~
f ~B
CO C-
II
ca S
P
«• CA
CQ V
S g
e i
Z 35
„
1
H
Interface
!••
D.S
U CS
a: a
—
•3
tn
a
•o
TT
•a
S
•a
S
o
S
CO
o
0
T3
to
"5»
i
VO
Tl-
1
tu
cu
3C
1
n:
i
2
u->
t
1
s
o
y
o
u
0
(2
b
B
•^
OS
V
•3
.—
D.
to
O
c
.
>-, >> C
I I Q
f -tr
0 0
ON O^
•^ "f f
1
i
(N
O
D-
a
Farrance direct shear device)
E
CO
1
£~
3= g
Sf
1 CO
fn o
>— ' J=
2*
•^ o
£?•?
ting Company, Inc., Canonsbui
Hants Ltd., Northbrook, IL (12
J&L = J<es
STS = STS Consu
•••*
\ (Wykeham Farrance device)
nch direct shear box)
°- '" ^
ity, Philadelphia,
g Laboratory (2A
direct shear box'
u .£
E c (N
ic Research Institute, Drexel U
of Texas at Austin, Civil Engii
nmental, Inc., Austin, Texas (1
GRI = Geosynthet
UTA = University
TRI = TRI Enviroi
reel shear box)
^
f
c
iciates, Denver, Colorado (12-i
1
"o
a
n
)
I shear box)
x u
r^ ai
ch direct shear b<
, PA (12-inch dir
.E "
(N ~
O c3
Consultants, Atlanta, Georgia
id Geotechnical Consultants, C
GSC = GeoSyntec
CGC = Cumberlar
irect shear box)
ID
•g
C
ineering, Manchester, NH (12-i
ME! = Miller Engi
X
o
ct shear box)
ch direct shear bi
.S -7
•5 tN
•= ;^-
>, Arlington Heights, IL (12-im
Terra Testing, Lakewood, CO
CETCO = CETCC
ATT = Advanced
^*s
i direct shear box
o
1
rs
V— '
0
CO
c
.8
0
S
AGP = AGP Labo
geotextile of Bentomat.
o
1
u
2
fN
-------
COLLOID ENVIRONMENTAL TECHNOLOGIES COMPANY
DIRECT SHEAR TESTING
1200
tr>
Q.
to
in
UJ
to
1100-
1000-
900 —
800
700 H
600
500-
400 —
300-
200-
100 —
TEST SERIES NUMBER 1: INTERNAL STRENGTH
OF HYDRATED BENTOMAT GCL
MEASURED SHEAR STRENGTHS
CTn = 50 psf, Tpeok = 340 psf
an = 150 psf, T^ = 520 psf
an = 300 psf, Tpeok = 765 psf
CTn = 500 psf, Teak = 1085
•••_•_• PEAK
-------
COLLOID ENVIRONMENTAL TECHNOLOGIES COMPANY
DIRECT SHEAR TESTING
1600
1400-
1200-
^ 1000-
UJ
o
O 800'
.
gj
CO
600-
400-
200-
TEST SERIES NUMBER 2: INTERNAL STRENGTH OF
DRY BENTOMAT GCL
TEST CONDITIONS
0.0
0.4
0.8
I I I I
1.2 1.6
I ' I ' I '
2.0 2.4 2.8 3.2
DISPLACEMENT (in.)
NOTE:-The shear box size was 12 in. by 12 in.(300 mm by 300 mm),
and the contact area remained constant throughout the
entire test.
DATE TESTED: 19 AUGUST 1993
JlSmtak. GEOSYNTEC CONSULTANTS
GEOMECHANICS AND ENVIRONMENTAL LABORATORY
FIGURE NO.
PROJECT. NO.
DOCUMENT NO.
1-3
GL3419
GEL93291
PAGE NO.
G-5
-------
COLLOID ENVIRONMENTAL TECHNOLOGIES COMPANY
DIRECT SHEAR TESTING
1600'
1400-
1200-
.2 1000
LU
O
cc.
O 800'
CO
600-
400-
200-
TEST SERIES NUMBER 1: INTERNAL STRENGTH OF
HYDRATED BENTOMAT GCL
TEST CONDITIONS
an = 50 psf
-------
ATTACHMENT 1
DIRECT SHEAR TEST RESULTS
G-7
-------
COLLOID ENVIRONMENTAL TECHNOLOGIES COMPANY
DIRECT SHEAR TESTING
24000
00
8
I
21000-
18000 -
15000-
12000-
9000-
6000 —
3000-
MEASURED SHEAR STRENGTHS
=20500 sf, T = 9845
705 psf
1720 psf
2840 psf
_• PEAK : $
_• RESIDUAL : /,
2£: °> = 700 psf; Rz - 0.996
7 : a, = 515 psf; Rz ~ 0.998
HYDRATION TIME: 48 hours
SHEAR RATE: 0.04 in./min
0 3000 6000 9000 12000 ' 15000 ' 18000 ' 21000 ' 240 0
NORMAL STRESS (psf)
m
CONSULTANTS
DATE TESTED: 10 TO 17 JANUARY
FIGURE NO. ,
GEOMECHANICS AND ENVIRONMENTAL LABORATORY
PROJECT NO.
DOCUMENT NO.
r.\
PAGE NO.
G-8
-------
COLLOID ENVIRONMENTAL TECHNOLOGIES COMPANY
DIRECT SHEAR TESTING
14000
UJ
O
or
e
or
x
00
13000 -
12000 —
11000 -
10000 —
9000 —
8000 -
7000-
6000 —
5000 -
4000-
3000 —
2000 -
1000-
0
TEST SERIES NUMBER 1: INTERNAL STRENGTH
OF HYDRATED BENTOMAT GCL
TEST CONDITIONS
an = 2000 psf
an =10000 psf
an =20500 psf
HYDRATION TIME: 48 hours
SHEAR RATE: 0.04 in./min.
0.0. 0.4 0.8 1.2 1.6 2.0
DISPLACEMENT (in.)
2.4
2.8
3.2
NOTE: The shear box size was 12 in. by 12 in.(300 mm by 300 mm).
and the contact area remained constant throughout the
entire test.
DATE TESTED: 10 TO 17 JANUARY 1994
jm^***. GEC>SYNTEC CONSULTANTS
GEOMECHANICS AND ENVIRONMENTAL LABORATORY
RGURE NO.
PROJECT NO.
DOCUMENT NO.
1-1
C,\ 3579
C5FI 94074
PAGE NO.
G-9
-------
ui
co
O£
UI ^_
b
in a? "*
51"
^SCD
COUJ °.
luces
a: i— g
L_ «X*
F"* 7T
agg
H* UJ ^
g^ 7fe «J
ilo^
loLU^
BC £
S&I
S_ic
*-««s S
QfeS
>-0
go
£-!
sS wj
" ^ ^^
«/» *••?
2°
ac
S
M
i
n2
*
*J
i1
0)
•M
S
QL
C£
g
"^ ij
_£ C
3
1.
If
u_
o:
i
*
O)
en
c
o
•4J
U
u.
111
o«~
1
O
A)
*
-M
•0
tt)
M J.
S 7! "§
0
c
c
in
ir
k
u
o
0
o
0
r**
•
CM
^
C
IT
C
CNJ
C
0
o
c
e*.
— j
4-*
3
c
cu
CO
T3
Ol
•o
=
u.
o
JC
*>
en
g
k
to
£
0)
4->
C
*~*
9-4
en
t.
fO
OJ
VI
o
fc.
o
n
O)
fl)
-M
CO
r—
P
O
C
TO
(U
I
O
re
ej
1
s
o
•^
oo
£
h
u
1
&
c
TJ
£
1
!3
X
1
lj
S.
V}
g
G)
JC
w
QJ
V
S
4-*
J
*J
o
iZ
5
(4
cn
shearin
u
o
T3
OJ
M
3
V)
Ol
U
n
"i
c
o
c
n-
o
c
o
Q.
D.
0)
4J
10
>»
0)
•5
1
•o
(U
i.
«
01
0)
i
•9
J
U
1
4J
CO
01
1=
£1
u -
o c
u- o
s^
"ie
t.
c o
0 0
M U-
01 O
.£=
13 4J
•0 C
01
ei --
4<* <«-
<4-
c o
H
C 4J
¥""
1°
M
0) >
e-t
*° 0)
2 ?E
g
•3 OT
OJ
C "^ •
U W 4-»
Q> W
"O 4-» 01
C •*-»
_C QJ
C -M J=
0) 13 •*->
s-g i
0}
— J > S"
CJ O *-
CO M g
** Ei ^
o t- o.
J-> _-.
je adhesion"
s range of s
ear strength
4-* ^M *"
** "w i
•S o ^
4-» CO
11?
S *- r—
§2 J
s i •*•
^ c "o
en c
"S^ S.
> J->
|o 3
B ^ -5
•o in "~
0) C C
4-1 O IO
1- -^
O 4-* 01
&S -o
f— ">
J & S
•—
-------
COLLOID ENVIRONMENTAL TECHNOLOGIES COMPANY
CREEP SHEAR TESTING
SHEARING PHASE
0.25
£, 0.20 -
\—
2 0.15 -
UJ
0.10 -
Q_
GO
5 0.05 -
o 0.
cr
> -0
-0
-0
00
.05 -
.10 -
.15
TEST SERIES NO. 1: INTERNAL STRENGTH OF
SOAKED BENTOMAT GCL
NORMAL STRESS: 500 psf
CONSTANT SHEAR LOAD: 250 Ibs
SHEAR DISPLACEMENT
VERTICAL DISPLACEMENT
SHEAR PHASE STARTED: 27 SEPTEMBER 1994
1 I
I
1 o -3 1 o -2 1 o
TIME (hours)
0 3 1 0 4
TESTING ATMOSPHERE: AIR MAINTAINED AND REGULARLY MON]TORED
AT A RELATIVE HUMIDITY OF 50 TO 72 PERCENT AND A TEMPERATURE
OF 70 +/- 2 degF(21 -f/~ 1
DATE REPORTED: 16 JUNE 1995
GEOSYMTEC CONSULTANTS
GEOMECHANICS AND ENVIRONMENTAL LABORATORY
FIGURE NO.
PROJECT NO.
GLI3545-02
DOCUMENT NO.
PAGE NO.
G-ll
-------
COLLOID ENVIRONMENTAL TECHNOLOGIES COMPANY
CREEP SHEAR TESTING
SHEARING PHASE
10 2u
10 i
1 -1
UJ
10-3i
10"S
co 10
Q
10
-7
TEST SERIES NO. 1: INTERNAL STRENGTH OF
SOAKED BENTOMAT GCL
NORMAL STRESS: 500 psf
CONSTANT SHEAR LOAD: 250 Ibs
SHEAR PHASE STARTED: 27 SEPTEMBER 1994
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14
SHEAR DISPLACEMENT (in.)
TESTING ATMOSPHERE: AIR MAINTAINED AND REGULARLY MONITORED
AT A RELATIVE HUMIDITY OF 50 TO 72 PERCENT AND A TEMPERATURE
OF 70 +/- 2 degF(21 -f/~ " ' ~"
DATE REPORTED: 16 JUNE 1995
jSS*mm^ GROSYNTEC CONSULTANTS
GEOMECHANICS AND ENVIRONMENTAL LABORATORY
FIGURE NO.
PROJECT NO.
GLI3545-02
DOCUMENT NO.
PAGE NO.
G-12
-------
COLLOID ENVIRONMENTAL TECHNOLOGIES COMPANY
CREEP SHEAR TESTING
SHEARING PHASE - LOAD INCREMENT'3
0.20
LJ
UJ
O
3
Q.
0.15 -
0.10 -
Q 0.05 -
5 o-oo H
p
C£
LJ
c -0.05 -
-0.10
TEST SERIES NO. 2: SOAKED BENTOMAT GCL /
80-m5l 6UNDLE TEXTURED HOPE GEOMEMBRANE
NORMAL STRESS: 6000 psf
CONSTANT SHEAR LOAD: 2100 Ibs
—~ SHEAR DISPL: UPPER SHEAR BOX
~*— SHEAR DISPL: LOWER GEOTEXTILE OF GCL
VERTICAL DISPL.
LOAD INCREMENT APPLIED: 27 MARCH 1995
10
i 11
-2
TTITTJ 1 I IHIH] TTTT
10"1 1
TIME (hours)
I I IHIII| I I llllll| TTTT
10 102 103 10
TESTING ATMOSPHERE: AIR MAINTAINED AND REGULARLY MONITORED
AT A RELATIVE HUMIDITY OF 50 TO 72 PERCENT AND A TEMPERATURE
OF 70 +/- 2 degF(21 +/- 1 degC).
DATE REPORTED: 16 JUNE 1995
JSSSSmL. GEC-SYNTHC CONSULTANTS
GEOMECHANICS AND ENVIRONMENTAL LABORATORY
FIGURE NO.
PROJECT NO. GLI3545
DOCUMENT NO.
PAGE NO.
G-13
-------
COLLOID ENVIRONMENTAL TECHNOLOGIES COMPANY
CREEP SHEAR TESTING
SHEARING PHASE - LOAD INCREMENT 3
1 1
10~3
10 ~4i
10
-5
-6
10 "°i
Ld
Q 10
~?J
10-S
10
TEST SERIES NO. 2: SOAKED BENTOMAT GCL /
80-mil GUNDLE TEXTURED HDPE GEOMEMBRANE
NORMAL STRESS: 6000 psf
CONSTANT SHEAR LOAD: 2100 Ibs
DISPL RATE: UPPER SHEAR BOX
DISPL RATE: LOWER GEOTEXTILE OF GCL
RELATIVE DISPL BETWEEN UPPER AND LOWER GEOTEXTILE
LOAD INCREMENT APPLIED: 27 MARCH 1995
-10. ^_^__^_^__^_^__^
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14
SHEAR DISPLACEMENT (in.)
TESTING ATMOSPHERE: AIR MAINTAINED AND REGULARLY MONITORED
AT A RELATIVE HUMIDITY OF 50 TO 72 PERCENT AND A TEMPERATURE
OF 70 +/- 2 degF(21 -f/- 1
DATE REPORTED: 16 JUNE 1995
GEOSYNTEC CONSULTANTS
GEOMECHANICS AND ENVIRONMENTAL LABORATORY
FIGURE NO.
PROJECT NO.
GLI3545-02
DOCUMENT NO.
PAGE NO.
G-14
-------
Appendix H
Summary of Bentofix Shear Test Results, Prepared by National Seal Co.
-------
-------
£8
Si!
05
C/)
1—
cd
(D
CO
oo
00
CM
CO
— x s
crs — ™
\W u_ —
c o i
u_ •*-•
® g
r- CD
O
CO
I I I
O
O
O
CO
O
O
O
CM
O
O
O
V)
0
.C
o
c
c
a>
E
-------
3000
Internal Shear Strength
Bentofix NS Thermal Lock
NSC Job* 85033P;Nor«al Stress - 720. 1440. & 2880 psl;Test Speed - 0.04 in/min
2000
CO
(0
0
i_
•*->
co
l_
cd
0
JZ
CO
1000
Peak Angle - 29°;Adhesion - 694 psf
Residual Angle - 4°;Adhesion - 130 psf
1000
2000
3000
Normal Stress (psf)
• Peak
Test Area - 12" x 12'
D Res.
T«sted 4/11, 12 & 13/95
H-2
-------
CM
05
Q.
co
2J.
to
TO
CD
.c
o
c
c
o
E
CD
O
JXJ
Q.
05
Q.
09
a
CO
X
00
I
(0
to
o
O
O
O
CO
O
o
o
to
o
o
o
o
o
o
CO
o
o
o
CJ
o
o
o
(sqi)
H-3
-------
200
Internal Shear Strength
Bentofix NW Thermal Lock
NSC Job* 8308BT;Normal Sires* - 140. 170. & 200 pii;Test Speed - 0.04 in/min
CO
O.
CO
CO
-------
NATIONAL SEAL COMPANY
CREEP SHEAR TESTING
SHEARING PHASE
0.15
0.10 -
LU
< 0.00
fe
UJ
o>-0.05 -
Ld
IE
CO
-0.10
TEST SERIES NO. 1: INTERNAL STRENGTH OF
BENTOFIX NS GCL (WITH WOVEN AND NONWOVEN
GEOTEXTILES)
NORMAL STRESS: 430 psf
CONSTANT SHEAR LOAD: 220 Ibs
* .*-•*--
SHEAR DISPLACEMENT
VERTICAL DISPLACEMENT
10
SHEAR PHASE STARTED: 6 DECEMBER 1994
Til—i i llllli|—l i i in
10
1 10 10
TIME (hours)
i i i uni[—i i 11
2 10 3 10 A
TESTING ATMOSPHERE: AIR MAINTAINED AND REGULARLY MONITORED
AT A RELATIVE HUMIDITY OF 50 TO 72 PERCENT AND A TEMPERATURE
OF 70 +/- 2 degF(21 -*•/- 1 degC).
DATE REPORTED: 28 APRIL 1995
jSSSm. GEOSYNTEC CONSULTANTS
GEOMECHANICS AND ENVIRONMENTAL LABORATORY
FIGURE NO.
PROJECTNO. HinS71-n7
DOCUMENT NO.
PAGE NO.
H-5
-------
0.15
0.10 -
Ld
LU
3 0.05 -|
CO
Q
< 0.00 -
P
cr
bJ
jS
CO
-0.10
NATIONAL SEAL COMPANY
CREEP SHEAR TESTING
SHEARING PHASE
TEST SERIES NO. 2: INTERNAL STRENGTH OF
BENTOFIX NW GCL (WITH NONWOVEN AND NONWOVEN
.GEOTEXTILES)
NORMAL STRESS: 430 psf
CONSTANT SHEAR LOAD: 220 Ibs
_ ,1...
SHEAR DISPLACEMENT
*•*•*-»•+ VERTICAL DISPLACEMENT
SHEAR PHASE STARTED: 6 DECEMBER 1994
1 0 ~2 10
"'
1 10 10 2
TIME (hours)
10
10 A
TESTING ATMOSPHERE: AIR MAINTAINED AND REGULARLY MONITORED
AT A RELATIVE HUMIDITY OF 50 TO 72 PERCENT AND A TEMPERATURE
OF 70 +/- 2 degF(21 +/- 1 degC).
DATE REPORTED: 28 APRIL 1995
GEOSYNTEC CONSULTANTS
GEOMECHANICS AND ENVIRONMENTAL LABORATORY
RGURE NO.
PROJECT NO.
HI I3571.-02
DOCUMENT NO.
PAGE NO.
H-6
-------
0.50
0.40 -
LoJ
o 0.30 -\
D_
CO
5 0.20 H
o
fe 0.10 -{
Ld
< o.oo -
Ld
CO
-0.10
NATIONAL SEAL COMPANY
CREEP SHEAR TESTING
SHEARING PHASE I
TEST SERIES NO. 3: INTERNAL STRENGTH OF
BENTOFIX NS GCL (WITH NONWOVEN AND NONWOVEN
GEOTEXTILES)
NORMAL STRESS: 60 psi
CONSTANT SHEAR LOAD: 4320 Ibs
SHEAR DISPLACEMENT
VERTICAL DISPLACEMENT
*_»..*.****
10
I i I
—2
10
"T-TT
1 10 10
TIME (hours)
i i i iini|—T-TT
2 1
i i n
1 0
TESTING ATMOSPHERE: AIR MAINTAINED AND REGULARLY MONITORED
AT A RELATIVE HUMIDITY OF 50 TO 72 PERCENT AND A TEMPERATURE
OF 70 +/- 2 degF(21 -f/~
HATF
IMF 1QQS TO 7 .HIIY 19Q5
GEOSYNTEC CONSULTANTS
GEOMECHANICS AND ENVIRONMENTAL LABORATORY
FIGURE NO.
PROJECT NO.
DOCUMENT NO.
CLI3571
PAGE NO.
H-7
-------
------- |