See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/245408104

Design of a landfill final cover system

Article in Geosynthetics International · January 2010

DOI: 10.1680/gein.2010.17.3.124

CITATIONS READS

3 45

2 authors:

Erik Newman Timothy D. Stark

AECOM University of Illinois, Urbana-Champaign
13 PUBLICATIONS 18 CITATIONS 204 PUBLICATIONS 2,450 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Track Support Inspection View project

Track Structure Transitions View project

All content following this page was uploaded by Timothy D. Stark on 18 August 2015.

The user has requested enhancement of the downloaded file.

 Geosynthetics International, 2010, 17, No.3

Design of a landfill final cover system

T. D. Stark] and E. 1. Newman 2
JProfessor 0/ Civil and Environmental Engineering, 2217 Newmark Civil Engineering Laboratory,

University o/JIlinois, 205 N. Mathews Ave, Urbana, lL 61801, USA, Telephone. + 1 217 3337394,

TetejclX: + 1 217 333-9464, E-mail: tstark@illinois.edu

2StaUt:ngin eet; URS Corporation, 1333 Broadway, Suite 800, Oakland, CA 94612. USA,

Telephone: + 1 5108743296, Tete/ax: + 1 5108743268, E-mail: erik_newman@urscorp. com

Received 7 August 2007, revised 26 January 2010, accepted 30 January 2010

ABSTRACT: This paper describes a final cover slope failure at a municipal solid waste
containment facility. The lessons learned from this case history include: (i) slope stabi lity analyses
should be conducted whenever field conditions difTer from initial design assumptions, such as a
steeper slope and different geosynthetics; (ii) publishcd values of interface strength/friction angle
should not be used for final des ig n, instead site-spccific interface testing should be used; (iii) final
cover slope angle should not exceed the lowest geosynthetic interface strength in the cover system
to prevent tension in the geosynthetics and/or progrcssive slope failure of the slope; and (v)
designers should resist the temptation to utilise a pre-existing final cover des ign without performing
the necessary field reconnai ssance, interface testing, analysis, and design for the new site.

KEYWORDS: Geosynthetics, Geosynthetic-lined slopes, Interface shear strength, Wa ste containment,

Strength, Stability, Shearbox test, Failure, Final cover system, Landfill

REFERENCE: Stark, T. D. & Newman, E. J. (20 I 0). Design of a landfill final cover systcm.
Geosynthetics [ntemational17 , No.3, 124-131. [doi: 1O.1680/gein.2010.17.3.124)

1. INTRODUCTION 2. INITIAL DESIGN AND

SPECIFICATIONS
This article describes a case history that illustrates a
number of important design, specification, bidding, and In this case, the initial final cover system was designed by
construction issues for landfill final cover systems. The a state agency. Thus, one state agency permitted or
case history involves an unlined landfill located on the approved the design of another state agency without any
east coast of the U11ited States. The landfill operated external review. This lack of external review can be
from 1977 to 1998 when an adjacent lined landfill was problematic when the design is constructed by a private
opened. The closed site consists of about 113 300 m 2 contractor because problems may develop with the design.
with a closed landfill area of about 89300 m 2 The If a state agency also performs the construction activities,
landfill site is flat and is underlain by sandy soils with external review may not be as critical because the state
groundwater near the ground surface. The closed can assume responsibility for the design errors instead of
portion of the landfill is unlined and does not have a dealing with a contractor claim.
leachate collection and removal system below the waste Figure I shows the original design which included two
and so the installation of a final cover system was unbonded nonwoven geotextiles, a 0.5-mm (20 mil) thick
pursued to reduce infiltration, leachate generation and polyvinyl chloride (PVC) geomembrane, and a single­
possible groundwater contamination. The applicable sided drainage composite. The single-sided drainage com­
state regulations require a minimum of 0.6 m of interim posite was a geonet with a nonwoven geotextile heat­
soil cover prior to final closure, the slopes to be bonded to the top side of the geonet. A double-sided
3H : I V or less, and a final cover system that contains drainage composite is a geonet with two nonwoven
a low permeability layer, namely a geomembrane, to geotextiles heat-bonded to the top and bottom sides of the
reduce infiltration. geonet. In this case, a single-sided drainage composite
This article reviews the initial design, specification, was used with an unbonded geotextile below the geonet
bidding, construction problems, and the remedial measures and not a double-sided drainage composite.
for this final cover system. The article also provides The unbonded geotextiles above and below the PVC
recommendations to help avoid some of the problems geomembrane wcre designed to provide a cushion between
encountered at this site in other future projects. the geomembrane and the overlying geonet and the under­

1072-6349 © 20 I 0 Thomas Telford Ltd 124

Design o( a land/ill /inal ('ova svslun 125
100 mm Topsoil
drainage composite. In addition, no interface testing or
analysis was performed on the geonet/PVC geomembrane
0,5 m Compacted soil interface created after removal of the un bonded geotextile.
Unfortunately the geonet/PVC geomembrane interface
m ..,:",:"..::-:..~.::: ..~.:-
- ..
Geonet with heat bonded geotextde above

- - - - - - - - - - 0,5 mm PVC geomembrane

2
270 Unhanded
nonwoven geotextile
eventually proved problematic as shown in Figure 2. In
summary, the stability analyses performed did not consider
site-specific geosynthetics in the cover system,
The final cover design tor this site utilized intertace
- 270 glm Unbonded nonwoven geotextile
strength parameters obtained from the available literature.
This is not advised because site conditions, geosynthetics,
0,6 m Existing interim soil cover and site soils usually differ from those lIsed to generate
the literature data. The importance of site-specific testing
can be illustrated by considering the PVC geomembrane
Waste interface strengths presented by Hillman and Stark (2001).
A designer might be tempted to use the strength values
Figure I. Original final cover system design with unbonded presented in Hillman and Stark (2001), as suggested by
geotextile below geonet Richardson and Chicca (2005), for the design of a site,
However, Hillman and Stark (2001) tested the faille side
of a 0.75-111111 thick PVC geomembrane and so their
lying interim soil cover. respectively. Fortunately, the interface strength values are only applicable to these
unbonded nonwoven geotextile underlying the geonet was materials and should not be used unless the site to be
removed from the design in the first addendum to the designed uti lizes the faille side of a 0.7 5-mm thick PVC
contract bid package. The inclusion of this unbonded geomembrane for the field interface. At this landfill, the
geotextile/geonet interface could have resulted in a weak smooth side of a 0.5-mm thick PVC geomembrane was
interface between the geonet and PVC geomembrane. This installed below the geonet rather than the faille side of a
observation is based on interface shear test results pre­ 0.75-mm thick PVC geomembrane, and so the Hillman
sented in Hillman and Stark (2001) for PVC geomem­ and Stark (2001) data is not applicable because the faille
brane/double-sided drainage composite interfaces. In this and smooth sides of a PVC geomembrane usually yield
testing, the nonwoven geotextile in contact with the different interface strength parameters and differences in
geomembrane delaminated leading to movement at the PVC geomembrane thickness also usually result in differ­
geotextile/geonet interface which exhibited a low shear ent interface strengths. For example, less embedment of
resistance. Therefore, an un bonded geotextile!geonet inter­ the geonet into the PVC geomembrane usually occurs with
face would have exhibited even lower shear resistance a thinner PVC geomembrane which results in a lower
than that reported by Hillman and Stark (2001). shear resistance. Thus, it is expected that the smooth side
The final cover system attempted by the contractor is of the 0.5-mm thick PVC geomembrane used at this site
shown in Figure 2 and did not include the unbonded would exhibit a different shear resistance to that of the
geotextile below the geonet but all of the other components faille side of the 0.7 5-mm thick membrane, as used by
from Figure I remained the same. Based on field observa­ Hillman and Stark (200 I) for the geonetlPVC geomem­
tions during construction, sliding occurred along the ge­ brane interface in Figure 2.
onet/PVC geomembrane interface shown in Figure 2. This cover design also resulted in the friction angle of
At this site no interface testing was conducted before, the weakest interface, namely the geonet/PVC geomem­
during, or after design to assess slope stability or the brane interface, being less than the final cover slope angle.
impact of an unbonded geotextile below the single-sided This will lead to tensile forces developing in the geosyn­
thetics and/or progressive slope failure of the final cover
200 mm Topsoil
slope even if the cover system is successfully completed.
It is recommended that the lowest interface friction angle
should exceed the steepest slope angle (Stark and Poeppel
0,5 m Compacted soil
1994; Gilbert and Byrne 1996; Stark and Choi 2004).
- - - - - - • Geonet with geotextile above

............................ ~ mm PVC geomembrane SLIDING

LOCATION 3. INITIAL CONSTRUCTION PROBLEMS
- - - - - - • 270 glm2 Nonwoven geotextile Before placement of the geosynthetics for the final cover
system commenced, the regulatory agency identified areas
0,6 m Existing final cover of the landfill that did not possess the specified depth of
0.6 m of interim soil cover. To remedy this deficiency the
owner placed additional interim soil cover on the existing
slopes. This caused some of the final slopes to be steeper
Waste
than the designed and regulated slope of 3H : I V Figure 3
Figure 2. Final cover system bid and attempted by is a plan view of the landfill which is shaped like a
contractor saucepan with a long handle. The slopes which were
GCO,lynlhctics International, 2010, 17, No.3
126 Stark and Newman

Location of
'.
roadway slide "

-- ". - - -- - .­

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

\
Areas with slopes steeper thatn 3H : 1V Area pictured in
\

Figure 4

Approximate extent of cover soil

placement at halt of construction ------
Approximate extent of geosynthetics
installation at halt of construction

--
o
Scale

60
metres

Each contour is 0.6 m

120
t
North

Figure 3. Plan view showing slopes steeper than 3H : IV after placement of additional soil to achieve required 0.6 m thickness of
intermediate soil cover

steeper than 3H: I V are primarily located along the

panhandle as shown in Figure 3. As a result, the field
conditions differed from the final design conditions with
the field slopes steeper than 3H : 1V
The slopes of the over-steepened areas could not be
remediated for the following two reasons. First, the slope
toe along the panhandle could not be moved outward to
reduce the slope inclination because adjacent properties
were within 0.3 to 1.6 m of the slope toe. The designer
sought a variance for the 0.6 m thickness of the interim
soil cover to reduce the slope inclination but the permit­
ting agency denied the request. This led to the placing of
additional cover soil by the owner and the contractor
trying to cover slopes steeper than 3H: I V Second, the Figure 4. Aerial view of panhandle and southwestern slope
contractor was not tasked with removing the recently
placed interim soil cover, removing waste to flatten the
slope to 3H : I V, disposing of the waste, and replacing the to place material down the western-most slope and then
interim soil cover to achieve a 3H : I V slope in the over­ install cover soil from the bottom to the top of the slope
steepened areas along the panhandle. In such a situation, along the north and south sides of the panhandle.
the stability analyses should be revised to reflect the over­ Before proceeding too far down the panhandle, a
steepened slopes (steeper than the 3H : I V design slope) portion of the roadway slid to the north off the narrowest
before the contractor is allowed to proceed to prevent portion of the panhandle crest (see arrow in Figure 4). The
problems from developing. If the stability analyses show slide movement clearly occurred at the geonet/PVC
that the over-steepened areas will not be stable then the geomembrane interface (see Figure 5) because the over­
design should be revised before the contractor is allowed lying single-sided drainage composite tore (see Figure 6)
to take action. and this left the PVC geomembrane exposed, and essen­
tially undamaged, on the slope. The drainage composite
tore because there was no constructed seam in the
4. FIRST SLOPE FAILURE drainage composite near the crest of the slope where the
The following paragraphs illustrate some of the problems tensile forces were the highest.
that developed when the contractor initiated construction Because a well-defined and unsupported slide block
of the final cover system shown in Figure 2. Access developed, this instability provided an ideal opportunity to
limitations at the north side of the panhandle prevented back-calculate the interface friction angle that was mobi­
the creation of an access road at the slope toe so the lized along the geonet/PVC geomembrane interface. The
contractor started placing vegetative cover soil over the small amount of tensile resistance of the geonet that was
installed geosynthetics on the crest of the panhandle from mobilized when the geonet tore was not considered in the
the east towards the west (see Figure 4). Upon reaching limit-equilibrium back-analysis of this slide described
the western end of the panhandle, the contractor intended subsequently but it was considered in a subsequent
Geo5y nthetics International, 20 I 0, 17, No.3
Design of a landfill final cover system 127

PVC geomembrane installed at this site using stockpiled

material to investigate the range of the interface strength
parameters for this interface after the failure. Each series
of interface tests were conducted using normal stresses of
4.8, 12.0 and 19.2 kPa to simulate the field normal
stresses and estimate the peak failure envelope. The peak
friction angle from these two series of tests varied from
14.4° to 18.4° and the adhesion/cohesion intercept for this
smooth interface varied from 1.5 to 0.2 kPa, respectively,
even though the same materials and laboratory were used
for each set of tests. The commercial laboratory reported
an adhesion for the geonet/PVC geomembrane interface at
normal stresses from 4.8 to 19.2 kPa even though the
smooth interface typically does not exhibit an adhesion as
discussed below.
Figure 5. Panhandle slide showing exposed surface of the Back-analysis of the panhandle instability was per­
PVC geomembrane formed in the present study using the computer program
XSTABL Version 5 (Shanna 1996) and the stability
method of Spencer (1967) as coded in XSTABL. Spencer's
stability method was used for the back-analysis because it
satisfies all conditions of equilibrium and provides the
best estimate of the factor of safety (Duncan and Wright
1980). The geometry for the cross-section used in the
back-analysis is shown in Figure 7 and was determined
from a site visit by the first author.
Table 1 presents the cover soil properties used in the
back analysis. These properties were obtained from a
report prepared by a consulting engineering company
hired by the state. The adhesion of the geonet/PVC
geomembrane interface was assigned a value of zero in
the back-analysis because this is a smooth interface that
does not exhibit a 'Velcro' effect. Thus, the friction angle
of the geonet/PVC geomembrane interface was back­
calculated; that is, varied using XSTABL until a factor of
safety of unity was obtained.
Figure 6. Panhandle slide showing torn single-sided drainage
The inclination of the slope in the panhandle instability
composite (see arrow) and intact geomembrane
area was approximately 2.74H : 1V (20.0°) based on the
topographic survey performed by the contractor after
back-analysis using the stability methodology presented
by Giroud et al. (1995).
Back-analysis of slope failures, especially those with 6
simple geometries such as the panhandle slide, provides a 6.1 m
good estimate of the mobilized or field interface strength IQ)
u
(Stark and Eid 1998). Laboratory shear tests try to c
ro
simulate field conditions but for a number of reasons u; 3
'6 Cover soil

laboratory shear tests are only an approximation of the rou (/I = 30.2°

:0:;

field conditions (Stark et al. 2000). In addition, problems Q; c = 5.7 kPa

> y = 19.9 kN/m 3

with the interpretation of laboratory test results, which

will be discussed subsequently, are alleviated by using a 0
0 3 6 9 12
back-analysis to estimate the field interface strength. For Horizontal distance (m)
example, two series of interface shear tests were con­
ducted by a commercial testing laboratory on the geonetl Figure 7. Geometry used in back-analysis of panhandle slide

Table 1. Cover soil and geonet/PVC geomembrane interface parameters used in back-analysis

Material Unit weight (kN /m 3 ) Cohesion/adhesion (kPa) Friction angle C)

Cover Soil 19.9 5.7 30.2

Geonet/ PYC geomembrane interface N/A 0 16.0 (back-calculated)

Geosynthetics international, 2010, 17, NO.3

128 Stark and Newman

placement of the additional cover soil by the owner (see This allows the contribution of various factors in the
Figure 3). This slope angle yielded a back-calculated geosynthetic-soil layered system to be identified and
interface friction of 16.0° after applying a 5% quantified. For example, the first two tenns of the equa­
reduction to correct for the three-dimensional effects of tion represent the friction and adhesion strength para­
the cover soil along the sides of the slide block (Stark and meters of the geosynthetic interface. The next two terms
Eid 1998). The back-calculated friction angle of 16.0° is of the equation represent the shear strength and geometry
in agreement with the range of friction angles (14.4 to parameters for the soil buttress at the slope toe. The last
18.4°) reported by the commercial laboratory. term in Equation represents the geosynthetic tension of
The adhesion value was set to zero in the back-analysis the geosynthetics in the slope.
because the geonet/PVC geomembrane interface does not In the back-analysis of the panhandle slide, no soil
exhibit an adhesion, which means that there is no shearing buttress was involved because the cover soil did not
resistance at zero normal stress or 'Velcro' effect. This extend to the slope toe. As a result, the two soil buttress
can be visualised by placing the geonet on a sample of the terms were removed from the factor of safety expression
PVC geomembrane and observing little to no resistance in Equation 1 as shown in Equation 2. Equation 1 was
when the geonet is slid along the surface of the geomem­ further simplified by removing the adhesion term because
brane with no normal stress applied. A non-zero value of the PVC geomembrane and interface does not
adhesion indicates that the interface exhibits a shear exhibit an adhesion as shown in Equation 2.
resistance at zero normal stress. An example of a geosyn­ tan 0 a t
thetic interface that exhibits an adhesion value greater than FS - - - + -------.:.--­
y t sin /3 h 8)
zero is a textured high-density polyethylene (HDPE) (1)
geomembrane/non-woven geotextile interface where the c cos ¢ T
+------'-----­
geotextile does not slide easily over the geomembrane y h sin /3 cos(/3 + 8) y ht
because of a 'Velcro' effect (Stark et al. 1996). The
inclusion of a non-zero adhesion for an interface subjected where a is the geosynthetic interface adhesion (kPa); 0 is
to low normal stress can be non-conservative because the the geosynthetic interface friction angle (0); c is the soil
component of shearing resistance derived from adhesion cohesion (kPa); ¢ is the soil friction angle CO); /3 is the
can be much larger than the component derived from the slope angle CO); t is the thickness of the soil cover (m); h is
friction which is proportional to the normal stress. the vertical height of the slope (m); y is the unit weight
The adhesion is not normal stress-dependent and is multi­ of the cover soil (kN/m 3); T is the geosynthetic tension
plied by the length of the failure surface at all nonnal (kN/m).
stresses. In summary, the back-analysis showed that the
geonet/PVC geomembrane interface installed at the site FS = tan 0 + T (2)
tan/3 yht
exhibited a friction angle of about 16.0° and an adhesion
of zero. Koerner (1998) an average peak tensile strength
The inclination of the slope in the panhandle instability of a bi-axial geonet in the machine direction (the geonet
area was approximately 2.74H : 1V or 20.0° which meant installed at this site is bi-axial and the geonet was oriented
that the slope angle exceeded the back-calculated interface with the machine direction aligned down the slope) of
friction angle (16°) of the geonet/PVC geomembrane 12.0 kN/m. Using Equation 2, a factor of safety of unity,
interface. This caused tensile stresses to develop in the slope geometry parameters in Figure 7, and the tensile
geosynthetics until they tore and the slope failed because resistance of the geonet from Koerner (1998), the interface
no toe buttress was present in this case. If a toe buttress friction angle, 0, can be back-calculated and compared
was the slope may have remained stable, at least with the XSTABL back-analysis using Equation 3:
for a short time, by relying on the toe buttressing.
However, progressive failure and the development of a
post-peak interface strength, namely an interface strength
o = lan- 1 (tantl y:tl) (3)

that is lower than the peak interface strength, has been Based on field observations, Figure 4 for example, the
shown to develop even with a toe buttress if the slope following values were used in Equation 4 to back­
angle is greater than the peak friction angle of the critical calculate the geonet/PVC geomembrane interface friction
interface (Stark and Poeppel 1994; Gilbert and angle: /3 20° (slope angle) (2.74H: 1V); t 1.0 m ;
1996; Stark and Choi 2004). As a result, the weakest h = 3.1 m (for a slope length of 6.1 m and slope angle of
interface should exhibit an interface friction angle that 20°, see Figure 7); y 18.1 kN/m 3 ; T = 12.0 kN/m :
exceeds the slope angle to prevent the occurrence of shear
displacement on the weakest interface, development of a
post-peak strength along the interface, and progressive
12.0
() = tan - 1 ( tan(200) [ 1 ------:--:------­
(19.9 )
failure. = 14.5°
Giroud et al. (1995) present a limit equilibrium method
(4)
to evaluate the stability of geosynthetic-soil layered
systems of slopes. The main benefit of this method is that The XSTABL back-analysis yielded a back-calculated
the factor of safety is expressed by an equation that friction angle of about 16° so the geonet tensile strength
consists of the sum of the five terms shown in Equation 1. may have contributed the equivalent of about 1.5° of shear
Geosynthetics international, 2010, 17, No.3
Design of a landfill final cover .system 129
12.5
resistance to this geonet/PVC geomembrane interface
because of the low normal stress applied. Given the slope 10 .
angle was about 20° (2.74H : 1V) in this area, the m
0..
6
was probably placed in tension shortly after the cover soil if)
if)
7.5
was applied to the slope crest because the slope ~
(j)
5.0
exceeded the interface friction angle. The geonet/PVC Co
Cl.J
..c:
geomembrane interface started to deform until the geonet (f)
2.5
tore and the slide occurred shortly thereafter as evidenced
by the road only having advanced slightly past the slide 0
0 5 10 15 20 25
area (see Figure 4). Accepted design practice is for Normal stress (kPa)
geosynthetics to be designed without tension, that is,
having a friction that exceeds the weakest interface. Figure 9. Peak and residual torsional ring shear failure
In summary, the geonet/PVC geomembrane interface in envelopes for geonet/PVC geomembrane interface
the panhandle slide area exhibited a mobilized friction
angle of 15 to 16° and an adhesion of zero. This interface
friction angle was insufficient to support the 1.0-m thick envelope, because a ring shear device was used and a
of cover soil on the north slope of the panhandle and shear displacement of about 110 mm was achieved to
resulted in the first slope instability at the site. reach a residual strength condition.
The authors conducted site-specific interface testing of The peak and residual failure envelopes were stress
the geonet/PVC geomembrane using the torsional ring dependent and corresponded to peak and residual friction
shear test procedure described in Hillman and Stark angles of approximately 18 and 13°, respectively. The peak
(200 I). Three shear tests were conducted using stock­ strength increased in a non-linear fashion at a normal
piled geosynthetics from the site and normal stresses of stress of 20 kPa because the geonet started to embed into
4.8, 12.0 and 19.2 kPa to simulate the 0.6 m of cover soil. the PVC geomembrane. These friction are also less
The ring shear tests were conducted at a shear displace­ than the slope angles of 18.4° (3H:IV) to 21.8°
ment rate of 0.37 mm/min and the specimens were not (2.5H : 1V) observed at the site and confirm that instabil­
submerged. The geomembrane and geotextile of the ity should have occurred. The measured peak friction
single-sided drainage composite were glued to a Lucite angle is also in agreement with the back-calculated
ring using a thin coat of epoxy cement and allowed to friction of 16° and range of friction angles measured
cure separately for 24 h under a normal stress of approxi­ by the commercial laboratory (14.4 to 18.4°) with the
mately 19.2 kPa to create the geonet/PVC geomembrane. observed difference being caused by differences in field
After curing of the epoxy cement, the geomembrane and and laboratory conditions.
drainage composite were placed in contact with each other
in the ring shear device. The specimen container and
geosynthetics were marked to ensure that the geosyn­
5. SECOND SLOPE FAILURE
thetics did not slip during shear. The ring shear tests were In an effort to keep the project progressing while the
conducted at a laboratory temperature of 20°e. designer evaluated the panhandle slide, the contractor
Figure 8 presents the shear stress-displacement rela­ moved to the southernmost portion of the landfill, namely
tionships from the 0.5 mm PVC geomembrane/single­ the bottom of the 'pan' (see dashed box in Figure 4),
sided drainage composite ring shear interface tests. Using where installation of the geosynthetics had been recently
these relationships, the peak and residual failure envelopes completed (geosynthetics were still being installed in the
for the geonet/PVC geomembrane interface were con­ north-east comer, i.e. the top of the 'pan', at this time).
structed and are shown in Figure 9. A residual failure The contractor started placing cover soil from the
envelope is also plotted, not a large displacement failure bottom to the top of the south slope. As fill was being
placed up the south slope, tension started developing in
8
the single-sided drainage composite at the top of the
landfill. This tension manifested itself by failure of some
7
of the plastic ties holding the panels of the drainage
composite together and exposing the PVC geomembrane.
As a result, soil placement did not reach the top of the
south slope before the drainage composite started pulling
apart (see box in Figure 4) and sliding off the top of the
landfill.
Upon observing the drainage composite sliding off the
top of the landfill, the contractor concluded the geonet/
o~~~~~~~~~~~~~~~~~~~~~
PVC geomembrane interface was inadequate to support
o 0.02 0.04 0.06 0.08 0.1 0.12 0.14
Shear displacement (m)
the placement of cover soil and ceased work. Eventually
the owner terminated the contract so that the state could
Figure 8. Torsional ring shear test results on geonet/PVC redesign the cover system. The redesigned cover system
geomembrane interface was completed at an additional cost of US$l.4 million
Geo5ynthetics International, 2010, 17, No.3
130 Stark and Newman

over the contract price of USS2.0 million. The redesigned should be conducted on the potentially weak
cover system utilized a textured HDPE geomembrane and interfaces prior to bidding to confirm that the
a double-sided drainage composite to develop a Velcro required interface strength can be achieved with
effect with the geomembrane. The cover was constructed commercially available products.
using a different contractor. The first contractor was paid • The final cover slope angle should not exceed the
for all work done but unfortunately its termination was not lowest geosynthetic interface strength in the cover
a good business experience for the contractor even though system because progressive failure can occur along
it was not involved with the original design. that interface and lead to slope instability. At a
mmlmum this condition can lead to tension
developing in the geosynthetics, damage to the
6. PERFORMANCE OF PVC
geosynthetics, shear displacement occurring along
GEOMEMBRANES
the weak interface, and/or development of a post­
The good news from this case history is the perfonnance of peak strength.
the 0.5-mm thick PVC geomembrane. After 24 months • External review should be required for designs that
of ultra-violet (UV) light exposure in the period after most are created and approved by different state agencies
of the single-sided drainage composite had slid off the if the design is to be submitted for bid and
slope, the 0.5-mm thick PVC geomembrane still met all of construction by private contractors. If a state agency
the original project specifications and did not require will perform the construction activities, external
removal and replacement. The exposed PVC geomembrane review may not be as critical because the state can
was tested for volatile loss (ASTM D12(3), dimensional assume responsibility for design errors.
stability (ASTM D 12(4), puncture resistance (ASTM
D4833), hydrostatic resistance (ASTM D751), brittleness
(ASTM DI790), and basic tensile properties of plastic
sheeting (ASTM D882 Method A) and the test results NOTATIONS
exceeded the PVC Geomembrane Institute (2004) 1104 Basic SI units are given in parentheses.
specification for new PVC geomembranes. This is in
agreement with other case histories showing the long-term a geosynthetic interface adhesion (Pa)
durability of PVC geomembranes in a variety of environ­ () geosynthetic interface friction angle CO)
ments (Stark et al. 2001, 2(05). Conversely, all of the c soil cohesion (Pa)
single-sided drainage composite was damaged by UV light o soil friction angle CO)
exposure over the 24 months that the material was uncov­ f3 slope angle CO)
ered and had to be removed from portions of the cover that t thickness of soil cover (m)
did not fail due to cover soil placement. h vertical height of slope (m)
y unit weight of cover soil (N/m 3 )
7. LESSONS LEARNED AND T geosynthetic tension (N/m).
RECOMMENDATIONS
The technical and construction-related lessons and recom­
mendations generated from this challenging project in­
REFERENCES
clude the following suggestions.
ASTM D751. Standard Test Method for Coated Fabrics. ASTM
International, West Conshohocken, PA, USA.
• Design entities should not utilize prior designs
ASTM D882. Standard Test Method for Tensile Properties of Plastic
without performing the necessary field reconnais­ Sheeting. ASTM International, West Conshohocken. PA, USA.
sance, testing, analysis, and design for the new site. ASTM D1203. Standard Test Methodsfi)r Volatile Loss from Plastics
Engineering properties of geosynthetics and soils Using Activated Carbon Methods. ASTM International, West
vary from site to site and site-specific testing, Conshohocken, PA, USA.
ASTM D1204. Standard Test Methodfor Linear Dimensional Changes of
analysis, and design should be conducted.
Nonrigid Thermoplastic Sheeting or Film at Elevated Temperature.
• New slope stability analyses should be conducted ASTM International, West Conshohocken, PA, USA.
when field condi tions differ from the design ASTM D1790. Standard Test Method for Brittleness Temperature of
conditions, such as a change in geosynthetics or an Plastic Sheeting by Impact. ASTM International, West Conshohock­
increase in slope inclination, and/or length. In this en, PA, USA.
ASTM D4833. Standard Test Method for Index Puncture Resistance ol
case the stability analyses represent a 3H : 1V slope
Geomembranes and Related Products. ASTM International, West
with an unbonded nonwoven geotextile between the Conshohocken, PA, USA.
geomembrane and geonet but the actual slope Duncan, 1M. & Wright, S.G. (1980). The accuracy of equilibrium
inclination was as steep as 2.5H: 1V and the methods of slope stability analysis. In Infernational Symposium of
unbonded geotextile was removed leaving the weak Landslides, New Delhi, India, pp. 247-254 (also Engineering
geonet/PVC geomembrane interface. Geology, vol. 16, pp. 5-17,. Elsevier Scientific Publishing
Company, Amsterdam, The Netherlands)
• Published values of interface friction angle only Gilbert, R. B. & Byrne, R. J. (I 996). Strain-softening behavior of waste
should be used for preliminary Appropriate containment system interfaces. Geo,~YlIthetics International 3, No.
shear tests utilizing the actual materials involved 2,181203.

Geosynthetics International, 2010, 17, No.3

 Design of a landfill final cover .system 131

Giroud, J. P, Williams, N. D., Pelte, T. & Beech, .J. F. (1995). Stability of stability methods in practice. Journal of' Geotechnical and
geosynthetic-soil layered systems on slopes. Geosvnthetics Int(!l"na­ Geoenvironmental Engine(!l"ing, ASCE 124, No. II, 1049- 1060.
tional 2, No.6, 1115-1 148. Stark, T. D. & Choi, H. (2004). Peak versus residual interface strengths
Hillman, R. P & Stark, T. D. (200 I). Shear behavior of PVC for landfill liner and cover design. Gemynthetics Intel"llational 11,
geomembrane/geosynthetic interfaces. Ge()svnthetics Int(!l"national No. 6,491-498.
8, No.2, 135-162. Stark, T. D., Williamson, T. A. & Eid, H. T. (1996). HDPE
Koerner, R. M. (1998). Designing with Geo~Yflthetics, 4th edition. geomembrane/geotextile interface shear strength. Journal of" Geo­
Prentice Hall, Upper Saddle River, NJ, USA. technical Engineering 122, No.3, 197-203.
PVC Geomembrane Institute (2004). PVC Geomembrane Mat(!l"ial Stark, T. D., Eid, H. T., Evans, W. D. & Sherry, P. (2000). Municipal solid
Specification 1104. University of Illinois, Urbana, IL, USA. http:// waste landfill slope failure II: stability analyses. Journal of"
www.pvcgeomembrane.com (accessed I January 2004). Geotechnical and Geoeflvironmental Engineering, ASCE 126, No.
Richardson, G. R. & Chicca, W. E. (2005). Landfill closure: a lesson in 5,408-419.
crisis management. Geotechnical Fabrics Report 23, No.5, 18-21. Stark, T. D., Newman, E. J. & Rohe, F. P (2001). PVC Aquaculture liners
Sharma, S. (1996). XSTABL: An Integrated Slope Stability AJ1a~vsis stand the test of time. Geotec/mical Fabrics Report 19, No.7,
Program for P(!I"sonal ComfJuters. Interactive Software Designs, 16-19.
Inc., Moscow, Idaho, USA. Stark, T. D., Arellano, D., Horvath, J. & Leshchinsky, D. (2002).
Spencer, E. (1967). A method of analysis of the stability of embankments Guidelines for Geofoam Applications in Embankment Projects.
assuming parallel inter-slice forces. Geotechnique 17, No. I, Final report prepared for National Cooperative Highway Research
11-26. Program, Transportation Research Board, National Research
Stark, T. D. & Poeppel, A. R. (1994). Landfi II liner interface strengths Council, Washington, DC, USA.
from torsional ring shear tests. Journal of" Geotechnical Engineer­ Stark, T. D., Choi, H. & Diebel, P (2005). Influence of plasticizer
ing, ASCE 120, No.3, 597-615. molecular weight on plasticizer retention in PVC geomembranes.
Stark, T. D. & Eid, H. T. (1998). Performance of three-dimensional slope Geo~ynthetics International 12, No.2, 99-110.

The Editor welcomes discussion on all papers published in Geosynthetics International. Please email your contribution to
discussion@geosynthetics-international.com by 15 December 2010.

Ge05ynthetics international, 2010, 17, No.3

View publication stats