-
ilt
bul,FL, U
e dlowitsed sanlemstre
termed L-8 Storage Reservoirlls witsite isMarshArea 1ee Rivt of
thea Fedeon of frary pu
and provide protection against overtopping during hurricane
events,
Engineering Geology 151 (2012) 8999
Contents lists available at SciVerse ScienceDirect
Engineering
e lssupply to West Palm Beach.The reservoir is owned and
operated by the South Florida Water
Management District (SFWMD) and was previously mined by
PalmBeach Aggregates, Inc. (PBA) as a commercial limestone quarry.
The res-ervoir, shown in Fig. 1, was constructed during the period
20032008,consisting of seven interconnected cells that had
beenmined to approx-imately elevation 4.5 m North American Vertical
Datum, NAVD, (allelevations in this paper are NAVD-1988, unless
otherwise noted),
posits have relatively low permeabilities, which allow
utilization ofdry mining techniques for the quarry operation. These
characteristicsare also uniquely suited for water storage and
withdrawal with min-imal interaction with the surrounding ground
water system.According to current geologic literature, the shallow
geologic stratainclude variably lithied members referred to as the
Bermont Forma-tion (Petuch and Roberts, 2007). The upper-most
stratum at the site isidentied as the Fort Thompson Formation.
Overall, the ages of thethen the reservoir was further deepened b13
m to provide additional storage capa(Bromwell and Ying, 2010).
Perimeter e
Corresponding author. Tel.: +90 216 677 1630/197E-mail
addresses: [email protected] (A.T. zer)
(L.G. Bromwell).1 Tel.: +1 863 6672345; fax: +1 863 6672662.
0013-7952/$ see front matter 2012 Elsevier B.V.
Allhttp://dx.doi.org/10.1016/j.enggeo.2012.09.011rpose is to help
sustainals in order to facilitaterovide additional water
Bend geological formation, which underlies the entire L8SR, is
anarea of unique geological character. The shallow cemented
sandyshell of the carbonate-rich zone and the underlying shelly
sand de-desired ows and levels within adjacent canrestoration of
the Loxahatchee River and pThe Loxahatchee Reservoir, also(L8SR),
consists of 7 interconnected ceity of approximately 56106 m3.
TheBeach County, Florida, north of the Artuge (also termed Water
Conservationoriginal headwaters of the Loxahatch2009). The
reservoir is an integral parEverglades Restoration Plan
(CERP),prove water quality and the distributiida (Christmann et
al., 2009). Its primh a water storage capac-located in western
Palmall National Wildlife Ref-or WCA-1), within the
er (Bromwell and Ying,overall Comprehensiveral/State project to
im-esh water in south Flor-
as well as additional water storage (zer et al., 2009).
2. Site geology
Geologic conditions at the L8SR site include a surcial layer of
peatand organic sands underlain by sand intermixed with lenses of
silty toclayey sands, overlying a relatively thick deposit of
well-cementedsand and shell of marine origin (Land et al., 1973).
The 20-Mile1. Introduction elevation +7 m surround the reservoir.
The embankments wereconstructed from overburden excavated during
the mining process,Stability assessment of an earth dam on s
A. Tolga zer a,, Leslie G. Bromwell b,1
a Civil Engineering Department, Okan University, Tuzla Campus,
34959 Akrat/Tuzla, Istanb AMEC-BCI Engineers & Scientists Inc.,
2000 E. Edgewood Drive Ste.215, 33803 Lakeland,
a b s t r a c ta r t i c l e i n f o
Article history:Received 11 August 2011Received in revised form
12 September 2012Accepted 16 September 2012Available online 4
October 2012
Keywords:Rapid drawdownTransient seepageSlope stabilityCoupled
analysisSedimented silt/clay tailings
This case study describes thconsisting of low strength,within
the construction limCounty, Florida. Fully couplperformed for
operationalmethods (LEM) and nite etions utilizing fully coupledity
during drawdown.
j ourna l homepage: www.y dredging to elevationcity up to 56106
m3
arth embankments to
8; fax: +90 2166771486., [email protected]
rights reserved./clay tailings foundation: A case study
TurkeySA
esign, and construction of an earth dam built on recently
sedimented tailingsplasticity silt/clay, and variably silty sand.
The tailings were encountered
of Cell 7 at the 4 km2 Loxahatchee (L-8) Reservoir in western
Palm Beachteady state and transient seepage models with slope
stability analyses wered rapid drawdown conditions, respectively,
using both limit equilibriument stress-based modeling (FESBM)
techniques. In addition, stability calcula-ss/pore pressure
analysis were performed to analyze rapid drawdown stabil-
2012 Elsevier B.V. All rights reserved.
Geology
ev ie r .com/ locate /enggeoformations present at the PBA
facility range from approximatelytwo million years to less than
10,000 years.
3. Field exploration program
This paper concerns the design of the southern perimeter
em-bankment of Cell 7 using both limit equilibrium methods (LEM)
andnite element stress-based modeling (FESBM) methods to
provide
-
below the ground surface to the top of the rock formation.
Thesesoils ranged from 2.5 to 7 m in total thickness, depending on
theamount of overburden at the boring locations and the elevation
ofthe underlying rock formation. The surcial organic soils were
re-moved to facilitate mining operations, except at sporadic
locationswhere organic soils were covered by overburden. Limestone
and/orlimestone formations were encountered within the borings
underly-ing the surcial sands. These are the primary mineable units
encoun-tered at the site, and varied between 0.5 and 6 m in total
thickness.
Rock coring was performed to determine the quality of the
rockand to obtain representative samples. Recorded values for the
RockQuality Designation (RQD) ranged between 0 and 48 and
recoveryratio ranged between 17 and 83%. Blow counts (N-values)
recordedwithin these strata ranged from 24 blows per foot to
practical refusal(N-value>50 blows per 15 cm). Intermittent
layers of silty and clay-
Cells 1&2
Cell 3 Cell 4
Cell 5
Cell 7 Cell 6
90 A.T. zer, L.G. Bromwell / Engineering Geology 151 (2012)
8999adequate slope stability during normal operation as well as
rapiddrawdown conditions. Cell 7, as shown in Fig. 2, is
encompassed bythree exterior levees referred to as Cell 7 West,
South, and East. Cell7 encompasses a storage area of nearly 0.5
km2. It is bordered onthe south side by a 230 kV Florida Power
& Light (FP&L) transmissionline easement, the L-8 canal
along the east side, and the process plantalong the west side. A
conveyance structure in the form of an openchannel was dredged to
connect Cell 7 to Cell 6 through FP&L's trans-
NORTH
Fig. 1. Aerial view of Loxahatchee Reservoir.mission line
easement area. Cell 7 originally served as a processwater/tailings
disposal pond for limestone quarry operations, whichpreceded
reservoir construction at the site.
A eld exploration program was performed to dene the general-ized
site-specic soil and rock stratigraphy in order to evaluate
theexisting foundation conditions for design of the reservoir
embank-ments. A combination of mechanical auger and standard
penetrationtest (SPT) borings was completed to depths ranging
between 3.5 and20 m below the existing grade using tracked and
pneumatic drill rigs.
In the vicinity of the southern embankment of Cell 7, seven
soilstrata generally characterize the subsurface geotechnical
prole, asshown in Fig. 3. Typically, the surcial soils consist of a
combinationof interbedded layers of variably silty and clayey sands
extending
incremental loading (IL) consolidation tests were performed
on
L-8 CAN
PROCESS WATER/TAILINGS DISPOSAL POND (CELL 7)
PROCESS PLANT NORTH
CELL 7 SOUTH CELL 7 WEST
CELL 7 EAST
Fig. 2. Aerial vieundisturbed silt/clay tailings extracted from
the southwestern bermof Cell 7 to provide soil properties for
settlement analyses. The resultsof these tests are presented in
Table 4.
Based on the soil types encountered, the southwestern Cell 7
typ-ical cross section was divided into layers as follows:
embankment ll,sandy ll material, very loose sand, silt/clay
tailings, silty sands, andunderlying sands with variable quantities
of shell and shell fragments(Figure 3). In-situ and laboratory
permeability tests were performed
AL
CONVEYANCE STRUCTURE
CELL 6 ey sands were encountered within the rock. In addition,
within thefootprint of the southwestern embankment of Cell 7, low
plasticityclay, silt, and variably silty sand tailings, which
originate from theplant washing and sizing operations, were
encountered. Relativelylow N-values were recorded within the low
plasticity clay and silttailings, ranging from weight-of-rod to 10
blows per foot. N-valuesbetween 1 and 50 blows per foot were
recorded within the variablysilty sand tailings.
4. In-situ and laboratory testing
Both in-situ and laboratory tests were performed to determine
hy-draulic conductivity and strength properties of the materials to
use inLEM and FESBMmodels. Laboratory strength tests were conducted
onselected samples of embankment ll and silt/clay tailings.
Triaxialshear tests (TX) were performed on representative grab
samples ofoverburden used to construct the perimeter embankment
(BCI,2007; zer et al., 2009). The soil samples were compacted in
the lab-oratory to 95% of Modied Proctor maximum density, which was
theminimum compaction criteria for the as-built embankments. The
re-sults of the TX testing are presented in Table 1.
In addition, three TX tests, one direct shear (DS) test, and one
di-rect simple shear (DSS) test were performed on selected
undisturbedsilt/clay tailing samples extracted from the
southwestern berm of Cell7. The results of these tests are
presented in Tables 2 and 3. Also, fourw of Cell 7.
-
tions for different types of soils built into the software
(Geo-Slope
1 Embankment Fill 4 Silt/Clay Tailings2 Sandy Fill Material 5
Silty Sand3 Very Loose Sand 6 Sand with Shell
9
0
-9
-18
12
3 45
6
Foundation Sand with Shell I
Foundation Sand with Shell II
Bench El. -4.5 m 10H:1V
4H:1V
Crest El. 7 m Crest Width: Min. 4 m, Max 8 m.
Normal Operating Pool Level El. 4.5
nce
tion
(mete
rs, N
AVD)
15 m
of
91A.T. zer, L.G. Bromwell / Engineering Geology 151 (2012)
8999for seepage modeling. Laboratory permeability tests were
conductedon representative soils used to construct the southern
perimeter em-bankment of Cell 7. The soil samples were compacted in
the laborato-ry to 95% of Modied Proctor maximum density. In
addition, in-situpump well tests were conducted to estimate the
permeabilities ofthe subsurface soils (sandy ll material, very
loose sand, silty sand,sand with shell, and foundation sands with
shell) during the eld in-vestigations. The horizontal permeability
and ratio of vertical to hor-izontal permeabilities were calculated
based on data obtained fromin-situ and laboratory tests. The soil
layer permeability values areshown in Table 5. These values were
used in the steady state andtransient seepage analyses.
5. Seepage and stability analyses methods used
According to Lambe and Silva (2003), the components involved
indetermining an accurate factor of safety for earth structures are
situ-ation (events), geometry (composition), stress and stress
path, porepressure, strength, level of stability and probability of
failure. Withthe introduction of advanced software products along
with increasedcomputing power, powerful techniques such as nite
element com-puted stresses have gained popularity in geotechnical
engineeringdesign. In this design study for the southwestern
embankment ofCell 7, coupled steady-state and transient seepage and
deformationanalyses using both LEM and FESBM methods were performed
inorder to design the earth dam on silt/clay tailings, taking into
accountthe components listed by Lambe and Silva (2003).
5.1. Seepage analyses
Steady-state seepage analyses and transient seepage analyseswere
made in order to predict the location of the phreatic surface
-27 6030
Dista
Elev
a
0
Fig. 3. Original design cross-sectionand pore water pressures
for steady state and drawdown conditions,respectively. The typical
cross-section of the Cell 7 southwestern em-bankment was analyzed,
using the nite element program SEEP/W(Geo-Slope International Ltd.,
2008a) to determine the ow throughthe embankment based on the
embankment geometry, boundarywater pressures, soil prole, and
vertical and horizontal soil perme-abilities (hydraulic
conductivities).
Table 1Multi-stage triaxial (TX) test results.
Grab sample ID Effective stress envelope
Cohesionc '(kPa)
Friction angle '(degrees)
White sand 0 36Light gray sand 0 38International Ltd., 2008a),
and two closed form equations based oncurve t parameters (Van
Genuchten, 1980 and Fredlund and Xing,1994). Empirical relations
based on the grain size distribution forsandy soils presented by
Yang et al. (2004) were used in this studyto estimate Fredlund and
Xing (1994) curve tting parameters to de-ne the water
characteristics curve for sandy soils within the prole.
The Fredlund and Xing (1994) curve tting parameters for
sandysoils are summarized in Table 6, and water retention curves
areshown in Fig. 4. The retention curve parameters for foundation
sandlayers are irrelevant for the drawdown models since these
layers re-main always saturated.
Since the permeability of the silt/clay tailings are very
low(Table 5) unsaturated phenomena play a signicant role in their
be-havior (Zandarin et al., 2009). The Aubertin et al. (2003)
methodwas used to determine the water characteristic curve for
silt/clay tail-ings, since it can be applied for a variety of
materials, including tail-ings (Geo-Slope International Ltd.,
2008a). The resulting curve isalso shown in Fig. 4.
5.2. Stability analysis
Critical loading conditions that may occur over the operational
lifeof the structure were analyzed in order to determine the
minimumcalculated factor of safety (FOS). These loading conditions
and thecorresponding analyses are commonly designated as the
steady-Partially and fully saturated conditions can be modeled
usingSEEP/W. Volumetric water content functions for the drainage of
theupstream slope during drawdown were determined for the
unsatu-rated zone above the phreatic surface. In order to dene
volumetricwater content functions, SEEP/W offers four methods: one
based ongrain size (Aubertin et al., 2003), one based on a sample
set of func-
120 15090 180
(meters)
Cell 7 Southwestern Embankment.state seepage and rapid drawdown
cases. Several different analy-ses are available to determine the
FOS of the embankments, includingLEM (method of slices), boundary
element methods (Jiang, 1990),and nite element computed stresses
(Matsui and San, 1992;Dawson et al., 1999; Grifts and Lane, 1999;
Belirgen, 2007; Krahn,2007; Alkasawneh et al., 2008, and Huang and
Jia, 2009). In thisstudy both LEM and FESBM were used for the
design of the
Table 2Multi-stage triaxial (TX) and DS test results for
southwestern embankment of Cell 7.
Soil Effective stress envelope
Cohesion c ' (kPa) Friction angle ' (degrees)
Silt/clay tailings (TX test#1) 15 40.3Variably silty sand
tailings (TX) 17 39.3Silt/clay tailings (TX test#2) 9 36.5Silt/clay
tailings (DS) 6 32.0
alban.kuriqiHighlight
alban.kuriqiHighlight
-
southwestern embankment of Cell 7 for both steady state and
draw-down conditions.
LEM is based on the method of slices, and the FOS is dened as
theratio between available and mobilized shear strength along the
criti-cal failure plane. LEM, which is based only on the principles
of statics,has long been used in geotechnical practice for its
simplicity com-pared to other methods. However, Krahn (2003)
indicated thatsince the LEM does not deal with strain and
displacements it doesnot satisfy displacement compatibility. Krahn
(2003) recommendeda stressstrain based approach using nite element
computed stress-es (FESBM) inside a limit equilibrium framework. In
this method, -nite element computed stresses are rst calculated
then they areimported into conventional limit equilibrium
analysis.
On the other hand, Duncan (1996) indicated that accurate
LEMmethods of slices give the same values of FOS as obtained from
frictioncircle analyses, log spiral analyses, and nite element
analyses. Duncan(1996) also reported that themaximum difference
between FOS valuescalculated by methods that satisfy limiting
equilibrium conditions is
properties and the cost of nite-element analyses are high
(Duncan,1996).
In this design study, slope stability analyses were performed
forthe southwestern embankment of Cell 7 using the computer
programSLOPE/W (Geo-Slope International Ltd., 2008b). SLOPE/W
imple-ments various LEM (including General Limiting Equilibrium
Method,Ordinary or Fellenius Method, Bishop's simplied Method,
Janbu'ssimplied Method, Spencer Method, MorgensenPrice Method,Corps
of Engineers Method, LoweKaraath Method, Sarma Method,and Janbu's
generalized Method). For LEM practice, Krahn (2003)recommended
using Morgenstern and Price (1967) or Spencer(1967) methods since
these methods satisfy both force and momentequilibrium. Also, the
Spencer (1967) method was recommended byWright (1969) after
examining eight different LEM. In this study, allthe reported FOS
values using LEM are calculated using the Spencermethod. LEM
steady-state slope stability was investigated using Ef-fective
Stress Analyses (ESA). ESA uses predicted pore pressures and
Table 3DSS test results for southwestern embankment of Cell
7.
Soil Parameter Test 1 Test 2
Silt/clay tailings Vertical consolidation stress (kPa) 190
380Shear strain at the peak shear stress (%) 9.9 16.6Corrected peak
shear stress (kPa) 46 97Su 0vc
0.24 0.25
dex
Horizontal (kh) Vertical (kv)
Table 6Fredlund and Xing (1994) tting parameters for sandy
soils.
Soil a(kPa)
n m
Embankment ll (compacted sand) 1.74 4.47 0.62Sandy ll material
1.77 5.04 0.67Very loose sand layer 1.65 4.71 0.71Silty sand 2.59
4.39 0.59Sand with shell 1.67 5.37 0.66
92 A.T. zer, L.G. Bromwell / Engineering Geology 151 (2012)
8999about 12%, and concluded that an accuracy of about 6% for
calculatedFOS values is close enough for practical purposes.
Performing stability analyses using FESBM based on nite
elementcomputed stresses inside a limit equilibrium frame work
allowsconsideration of soil-structure interaction effects (Krahn,
2003). Inaddition, by calculating stresses using nite element
techniques, thereis no need to make assumptions about the
interslice forces (Krahn,2003). Using nite elementmodeling it is
possible tomodel many com-plex conditions, including non-linear
stressstrain behavior, non-homogeneous conditions, and changes in
the geometry; however,both the efforts involved in developing the
appropriate engineering
Table 4IL consolidation test results.
Soil Void ratioeo
Compression inCc
Silt/clay tailings 0.785 0.159Silt/clay tailings 0.532
0.116Silt/clay tailings 0.492 0.081Silt/clay tailings 1.050
0.238
Table 5Soil permeability values used in seepage analyses.
Soil Elevation (meters)Embankment ll (compacted sand)a Crest (El
+7 m) to El +3 mSandy ll materialb El. +3 m to +1.5 mVery loose
sand layerb El. +1.5 m to 0.5 mSilt/clay tailingsc El. 0.5 m to 1.5
mSilty sandb El. 1.5 m to 3 mSand with shellb El. 3 m to 4.5
mFoundation sand with shell 1b El. 4.5 m to 15 mFoundation sand
with shell 2b El. 15 m to 26 ma Permeability values were from
laboratory triaxial test results.b Permeability values were based
on in-situ pump well test results.c Permeability values were back
calculated from IL consolidation test data.effective stress
strength parameters, ' and c '.In order to perform FESBM, the
stress-deformation computer pro-
gram SIGMA/W (Geo-Slope International Ltd., 2008c) was utilized
inconjunction with SLOPE/W. In order to calculate FOS of
ne-grainedsoils during rapid drawdown, consideration of stress
induced porepressures and their dissipation, as well as seepage
induced pore pres-sures are recommended (Belirgen, 2007). Fully
coupled nite ele-ment computed stress/pore-pressure analysis using
SIGMA/W alongwith SEEP/W and SLOPE/W provided FOS values during
rapid draw-down by taking stress induced pore-pressures into
account. In addi-tion, seepage induced rapid drawdown models were
also run.
Compression index(eld adjusted)Cc
Recompression indexCr
0.172 0.0090.124 0.00530.089 0.00280.270 0.0244
Layer no Permeability (m/s)1 7.0105 4.0105
2 3.0105 1.5105
3 3.0105 1.5105
4 1.2108 5.8109
5 2.4105 6.7106
6 1.4105 9.1106
7 1.8105 8.8106
8 1.1105 4.8106
alban.kuriqiHighlight
-
The linear-elastic model uses the general Hooke's Law, and
thusrelates strain increments to stress increments, and it affords
a system-atic method for relating modulus values to stresses
(Duncan, 1996).The main drawbacks of this model are that it cannot
consistently dis-tinguish between loading and unloading phases
(Schanz et al., 1999),and also doesn't model plastic deformations
in a fully logical way0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.1 1 10 100 1000
Vol
umet
ric W
ater
Con
tent
Suction (kPa)
Embankment FillSandy FillVery Loose SandSilt/Clay TailingsSilty
SandSand with Shell
Fig. 4. Soil-water characteristics curves.
Table 8Modied cam-clay parameters for silt/clay tailings.
Soil eo
Silt/clay tailings 1.050 0.270 0.0244
93A.T. zer, L.G. Bromwell / Engineering Geology 151 (2012)
8999In order to perform FESBM steady-state slope stability
analyses,the stressstrain relationships of the materials in the
model need tobe dened. SIGMA/W includes six different soil
constitutive modelsincluding the linear-elastic model, anisotropic
elastic model, non-linear elastic hyperbolic model, elasticplastic
model, and modiedCamClay model (Geo-Slope International Ltd.,
2008c).
The linear-elastic model is the simplest soil constitutive
model,and only requires Elastic Modulus (E) and Poisson's ratio
().Duncan (1996) pointed out that whenever the simplest
representa-tion of soil behavior would result in acceptable
accuracy, there is noneed to use a very complex constitutive model
to analyze the prob-lem. In addition, the linear-elastic model
always ensures a solutionsince there are no convergence difculties
as with non-linear consti-tutive models (Krahn, 2003).
The linear-elastic model was used for all the sandymaterials in
thecross-section, since the strains for the sandy materials are
expected tobe very small. In addition, it was considered that a
reasonable stressdistribution can be obtained using simple
linear-elastic model param-eters for sandy materials since there
are no zones of hard materials ora soft core in the southwestern
embankment of Cell 7.
In order to perform steady-state FESBM slope stability
calculationsfor tailings consisting of low strength, low plasticity
silt/clay, and var-iably silty sand, two other constitutive models,
namely the non-linearelastic hyperbolic and modied CamClay models,
were used in addi-tion to the linear-elastic model, and the results
were compared.SIGMA/W uses the formulation developed by Duncan and
Chang(1970) for the non-linear hyperbolic model. The advantage of
thismodel is that the required model parameters can be obtained
directlyfrom triaxial strength tests, which are commonly performed
to deter-mine engineering properties.Table 7Material properties for
both LEM and FEM analysis.
Material Unitweight, (kN/m3)
Internalfriction angle, ' (degrees)
Cohesion,c ' (kPa)
Poissonratio,
Elasticmodulus,E (MPa)
Embankment(compacted ll)a
18 36 0 0.35 56
Sandy llmaterialb
17 28 0 0.30 20
Very loose sandb 15 25 0 0.30 15Silt/clay tailingsa 14 36 9 0.40
15Silty sandb 17 29 0 0.35 20Sand with shellb 18 33 0 0.35
50Foundation sandwith shell 1b
20 35 0 0.35 80
Foundation sandwith shell 2b
20 38 0 0.35 100
a Based on laboratory triaxial test results.b Based on SPT
correlations.(Duncan, 1996). After considering the low compression
ratio, (0.06to 0.13), relatively low void ratio (0.5 to 1.0), and
also fairly lowand narrow range for the plasticity index (5 to 18),
the Duncan andChang hyperbolic model (Duncan and Chang, 1970) was
selected asone of the constitutive models for the FESBM stability
calculationsof silt/clay tailings.
The third constitutive model selected for silt/clay tailings was
themodied CamClay model, since relatively low SPT blow count
valueswere recorded within these tailings, ranging between
weight-of-rodand 10 blows. The modied CamClay model's formulation
inSIGMA/W is based on Britto and Gunn (1987). It is an effective
stressbased model and undrained analysis cannot be performed using
thismodel, which is considered a drawback of the model.
When performing rapid drawdown stability analyses based onFESBM,
the rst step was establishing the long-term seepage condi-tions
using SEEP/W. To establish stress/pore pressure analysis,in-situ
stress conditions have to be established using SIGMA/W. Thisphase
of the model uses pore pressure conditions from parentlong-term
seepage analyses, and the reservoir pressure was appliedas a
uid-pressure type of boundary condition along the upstreamface of
the reservoir. After establishing long-term seepage andin-situ
stress conditions, the SIGMA/W transient model was complet-ed by
simulating the removal of the pool water by applying twoboundary
conditions on the upstream face of the reservoir. They arethe
stressstrain uid pressure boundary function and the
hydraulicpressure boundary function. Then, nite element computed
stressesduring drawdown stage are brought into limit equilibrium
frame-work (SLOPE/W) to complete the analyses.
6. Seepage and stability analyses results
6.1. Steady state seepage and long-term stability analysis
results
Earth dams reach equilibrium after prolonged water storage
undernormal operational pool levels. It can take many years to
establishsteady state seepage through an earth dam. Steady state
stabilityanalysis does not take into account the time required to
reach equilib-rium. Steady state seepage analyses (long-term) were
made using thenite element program SEEP/W in order to predict the
equilibrium lo-cation of the phreatic surface within the
embankment. Reservoirlevels of 4.5 m (normal operating pool), 5 m
(maximum operating
Table 9Upstream static slope stability analysis with steady
state seepage.
Method Calculated FOS
Normaloperating pool
Maximumoperating pool
Maximumsurcharge poolLEM 2.91 3.05 3.26FESBMa 2.81 2.94
3.16FESBMb 2.80 2.93 3.15FESBMc 2.89 3.04 3.28
a With linear-elastic model for silt/clay tailings.b With
non-linear hyperbolic model for silt/clay tailings.c With modied
cam-clay model for silt/clay tailings.
alban.kuriqiHighlight
-
less than the variability in soil engineering properties. Thus,
LEM
1
34
2
5 6
Foundation Sand with Shell I
Foundation Sand with Shell II
1 Embankment Fill 4 Silt/Clay Tailings2 Sandy Fill Material 5
Silty Sand3 Very Loose Sand 6 Sand with Shell
Fig. 5. Steady-state critical failure plane determined by LEM
for operational pool (FOS=2.9).
13
42
5 6
Foundation Sand with Shell I
1 Embankment Fill 4 Silt/Clay Tailings2 Sandy Fill Material 5
Silty Sand3 Very Loose Sand 6 Sand with Shell
raw
94 A.T. zer, L.G. Bromwell / Engineering Geology 151 (2012)
8999pool), and 5.5 m (maximum surcharge pool) were used in
theanalyses.
The input parameters for the ESA (steady state) analyses are
sum-marized in Table 7. Shear strength values and elastic modulus
for theembankmentll and silt/clay tailing layerswere calculated
based on tri-axial tests (Tables 1 and 2). For the rest of
thematerials, conservative es-timates of the material properties
were made based on SPT data.
The input parameters for the modied CamClay model derivedfrom IL
consolidation test results were used for silt/clay tailings,
assummarized in Table 8. SIGMA/W allows the user to dene the
mod-ulus of elasticity versus stress curves for non-linear
hyperbolicmodels, thus the modulus versus stress curve from TX test
resultsshown in Table 2 are used for non-linear models of silt/clay
tailings.
The upstream slope stability with steady state seepage using
theow properties from Table 5 is shown in Table 9 for reservoir
levelsof 4.5 m (normal operating pool), 5 m (maximum operating
pool),and 5.5 m (maximum surcharge pool). Fig. 5 shows the critical
failureplane for the case of normal operating pool calculated using
LEM. Rel-atively close agreement was found between the critical
failure planes
Fig. 6. Phreatic line after 10 days of dobtained from both LEM
and the three FESBM methods for thesteady-state analysis.
Based on the results shown on Table 9, the upstream
embankmentslope for all reservoir levels exceeded the minimum
required FOS of1.5 for static slope stability under steady-state
seepage conditionsestablished by the United States Army Corps of
Engineers (USACE,2003). Regardless of the selected soil behavior
(linear-elastic,
1
1.1
1.2
1.3
1.4
1.5
0 2 4 6 8 10 12 14 16 18Fac
tor o
f Saf
ety
(Base
d on F
ESBM
)
Elapsed Time (Days)
Drawdown from 4.5 mDrawdown from 5.0 mDrawdown from 5.5 m
Fig. 7. Results of rapid drawdown analysis.analysis using a
method that satises both force and moment equilib-rium (i.e.,
Spencer's Method) performed equally as well as FESBMbased models
for this study. Among the FESBM models, thelinear-elastic model can
be used for low plasticity silt/clay tailingssince it provided
reasonable stress conditions and did not requiregreat effort to
obtain the input parameters relative to other morecomplex
constitutive models.
6.2. Rapid drawdown stability analysis results
Rapid drawdown is dened as the sudden lowering of the reser-voir
water level, and the rate of acceptable drawdown needs to be de-ned
for embankment dams. If the water within the embankmentnon-linear
hyperbolic, and modied CamClay) for the FESBMmodeling of silt/clay
tailings material, reasonably close agreementwas obtained with the
FOS value using Spencer's LEM method. Themaximum difference between
calculated FOS methods using LEMand FESBM are within approximately
+4 and 1%, which is much
Foundation Sand with Shell II
down from normal operating pool.cannot drain fast enough as the
reservoir is lowered, internal porepressure will build within the
dam and may cause the upstreamslope to fail. There are numerous
reported drawdown failures in theliterature including Pilarcitos
Dam, Walter Bouldin Dam, numerousriverbank slopes along Rio
Montario (Duncan et al., 1990), and bankslopes of Licking River
(Londono, 2004).
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Tota
l Hea
d (m
eters)
Elapsed Time (Days)
Transient Seepage ModelStress/Pore Pressure Coupled
ModelReservoir Pool
Fig. 8. Pore water pressure response.
-
In order to perform drawdown slope stability analyses,
transientseepage analyses were rst performed utilizing SEEP/W. In
order toperform transient seepage analyses for the rapid drawdown
condi-tion, steady state seepage analyses (long-term) were rst
performedutilizing SEEP/W in order to predict the equilibrium
location of thephreatic surface within the embankment. Once the
steady-state anal-ysis was completed, transient seepage modeling
was performed byapplying a drawdown rate of 0.3 m/day along the
upstream face ofthe embankment as a moving head boundary
condition.
Based on the results of the transient seepage analyses, ow
netswere generated for the southwestern embankment of Cell 7.
Thenthe ow nets were imported into conventional limit equilibrium
anal-ysis (SLOPE/W) to perform stability analyses according to the
USACE(2003) method. Fig. 6 shows the phreatic line after 10 days of
draw-down from normal operating pool level. A dot within the middle
ofthe silt/clay tailings layer in Fig. 6 is a reference point,
which will bediscussed further regarding the pore-pressure
response.
zer et al. (2009) reported the results of coupled transient
seep-age models with rapid drawdown slope stability analyses using
LEMin three stages in accordance with the procedures described
inUSACE (2003). The method was developed by Lowe and Karaath(1960),
and later modied by Wright and Duncan (1987), andDuncan et al.
(1990). zer et al. (2009) concluded that using the tri-axial shear
test results and the USACE (2003) procedure, the mini-mum required
factors of safety were satised (varied between 1.4and 1.6).
However, analyses using undrained strength values fromDSS testing
(Table 3) gave a calculated minimum factor of safetyvalues on the
order of 1.0 (varied between 1.0 and 1.03), which is un-acceptably
low for rapid drawdown.
In this study, coupled stress/pore pressure modeling results
fromrapid drawdown analyses were used in FESBM modeling in order
toinclude the effect of stress conditions and soil-structure
interactionduring the drawdown. The rapid drawdown transient
seepage analy-ses were performed for reservoir levels of 4.5 m
(normal operatingpool), 5 m (maximum operating pool), and 5.5 m
(maximum sur-charge pool). For the L8SR, the design drawdown rate
is 0.3 m/day.
Because linear-elastic analysis seemed adequate to obtain
reasonablestress conditions for silt/clay tailings during steady
state stability analy-ses, the linear-elastic soil properties were
also used for the FESBM anal-yses of rapid drawdown. As recommended
by zer et al. (2009)undrained DSS test results (Table 3) were used
for slope stability calcu-lations for the silt/clay tailings. As
pointed out by Ladd and DeGroot(2003), DSS tests better represent
typical stress conditions imposed ona soft foundation layer by an
overlying earth embankment. During DSStesting, the shear stresses
are applied horizontally, which is most likelythe case in the eld,
rather than the approximately 45 shear planethat occurs in triaxial
tests. In addition, theDSS test representsmobilizedshear strength
that would likely be the failure mechanism underundrained shear
conditions for embankments on soft clays. In addition,DSS test
results will generally show a lower strength value for soft
cohe-sive soils thanwill triaxial compression tests (Ladd and
DeGroot, 2003).
The results of transient stability models based on FESBM
areshown in Fig. 7. Analyses using the FESBM modeling indicated
mini-mum factor of safety values on the order of 1.0 (varied
between1.03 and 1.06) which is in close agreement with the LEM
valuesusing Spencer's method (factor of safety values varied
between 1.0and 1.03). For the rapid drawdown case, the maximum
difference be-tween calculated FOS methods using LEM and FESBM are
within
4
4.0
4.5
7
ng Reservoir Pool
FO
mid st
lace
oun
ion
(a)
b) Ti
95A.T. zer, L.G. Bromwell / Engineering Geology 151 (2012)
89990.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
0 1 2 3 4 5 6
Res
ervo
ir Po
ol E
leva
tion
duri
Dra
wdo
wn
(m)
This line corresponds to therapid drawdown for pumpe
6 m. wide toe buttress
9 m. wide over excavation and rep
1 Embankment Fill 4 Silt/Clay Tailings2 Sandy Fill Material 5
Silty Sand3 Very Loose Sand 6 Sand with Shell
F
Foundat
(ElapsedFig. 9. (a) Results of rapid drawdown analysis when toe
buttress and over-excava1
1.5
2
2.5
3
3.5
8 9 10 11 12 13 14 15
S
nimum required FOS = 1.3 against orage dams (USACE, 2003)
ment area1
34
2
56
dation Sand with Shell I
Sand with Shell II
Fact
or o
f Saf
ety
me (Days)tion are implemented. (b) Design cross-section and the
critical failure plane.
-
approximately 3%, which is, much less than the variability in
soil en-gineering properties. This indicates that the normal stress
distributionalong the slip surfaces are in good agreement in limit
equilibrium andnite element based models.
The generated excess pore water pressure behavior during
rapiddrawdown using transient seepage analysis and the fully
coupledstress/pore pressure model are compared within the silt/clay
layerat the location shown in Fig. 6. As shown in Fig. 8 both the
transientseepage and stress/pore pressure models produced nearly
identicalresults. Consequently, the coupled stress/pore pressure
model usinglinear-elastic material properties resulted in
reasonable values ofstress distribution during rapid drawdown.
Thus, LEM analysisusing a method that satises both force and moment
equilibrium
(i.e., Spencer's Method) performed equally as well as FESBM
basedmodels for this study.
Regardless of the model (FEM or FESBM) selected to analyze
therapid drawdown stability of the reservoir, the factor of safety
valuesagainst rapid drawdown were on the order of 1.0, which is
unaccept-ably low for the rapid drawdown condition. Therefore, the
design ofthe Cell 7 southwestern embankment called for the
over-excavationof very loose or soft deposits of mine tailings
encountered beneath theembankment footprint within 6 to 9 m of the
inside toe. In addition, a6-meter wide toe buttress was added from
the mine bottom at4.5 mup to an elevation of 0.5 m to provide
additional resistance againstslope instability in the event of
rapid drawdown of the reservoir andalso to create a bench for a
possible future roller compacted concrete
(a)
(b)
(c)
Shot Rock Placement at elevation -4.5 m
Embankment Construction
Natural Ground at elevation +6 m
Tailings Removal
Tailings Removal
96 A.T. zer, L.G. Bromwell / Engineering Geology 151 (2012)
8999Fig. 10. (a) Embankment ll placement; (b) tailings removal; and
(c) Cell 7 southern embankment in operation.
-
7ed T
e med
oun
ion
ace
(a
(b
97A.T. zer, L.G. Bromwell / Engineering Geology 151 (2012)
89990.0
0.51.0
1.52.0
2.53.0
3.54.0
4.5
0 1 2 3 4 5 6
Res
ervo
ir Po
ol E
leva
tion
durin
g D
raw
dow
n (m
)
Elaps
This line corresponds to thrapid drawdown for pump
1 Embankment Fill 4 Silt/Clay Tailings2 Sandy Fill Material 5
Silty Sand3 Very Loose Sand 6 Sand with Shell
F
Foundat
6 m. wide toe buttress
9 m. wide over excavation and replslope protection system. These
stabilization measures resulted in a dra-matic increase in the
calculated factor of safety against rapid drawdownfrom
approximately 1.0 to 2.8 (Figure 9).
7. Construction details and post-construction performance
The southwestern embankment design called for the
over-excavation of very loose or soft deposits of mine tailings
encounteredbeneath the embankment footprint within 6 to 9 m of the
inside toe.The extent of over-excavation depended on the location
of tailingsalong the alignment. Unsuitable tailings were replaced
with suitableembankment ll material to increase the factor of
safety of the slopeand to reduce potential post-construction
settlement. In addition, atoe buttress consisting of well-graded
limestone rockwas added duringconstruction to provide additional
resistance against slope instability inthe event of rapid drawdown
of the reservoir.
Vertical construction of the southern Cell 7 embankment
extend-ed from the pit bottom at Elevation4.5 m to the crest at
Elevation+7 m (Figure 10a). Prior to the start of construction, the
water levelin Cell 7 was lowered to Elevation5 m in order to
excavate tailingsmaterial and prepare the foundation for the
embankment. Thesouthern alignment of Cell 7 was constructed from
the pit bottom(Elevation 4.5 m) because of constraints imposed as a
conse-quence of original mining operations and the location of the
trans-mission line easement relative to the proposed
construction.
Cell 7, as shown in Fig. 2, is bordered on the south side by a
230 kVFP&L transmission easement, and a conveyance structure in
the formof an open channel was dredged to connect Cell 6 to Cell 7
throughthe transmission line easement area. Seepage into the south
side ofCell 7 from the adjacent Cell 6, which was at a higher water
level, re-quired installation of a steel sheetpile wall across the
conveyance
Fig. 11. (a) Results of rapid drawdown analysis for the
as-b1
1.5
2
2.5
3
3.5
8 9 10 11 12 13 14 15
Fact
or o
f Saf
ety
ime (Days)
Reservoir PoolFOS
inimum required FOS = 1.3 against storage dams (USACE, 2003)
1
46
4
4
5
dation Sand with Shell I
Sand with Shell II
6
ment area
)
)channel, which had been backlled in order to allow lowering of
thewater in Cell 7. The sheetpile wall reduced the seepage and
allowedexcavation of the tailings material.
After the pit bottom was exposed and dewatered by rim
ditching,placement of embankment ll began. In some areas along the
em-bankment, wet conditions were encountered in excavations
aftermine tailings had been removed. These excavations were
backlledwith well-graded compacted limestone rock to provide a
stable work-ing platform for the placement and compaction of
embankment ll(Figure 10a).
All unsuitable mine tailings were removed to the extent
practicaland replacedwith clean compactedll after the pit bottomwas
exposedin order to increase the factor of safety of the slope and
to reduce poten-tial post-construction settlement (Figure 10b).
Embankment ll wasthen placed by large earth-moving equipment and
spread by bulldozersin 30-cm thick lifts before compaction.Waterwas
added as necessary toachieve a moisture content near optimum, and
compaction wasperformed by large vibratory roller compactors. The
compacted surfaceof each lift was scaried before the overlying
layer was placed. The res-ervoir was further deepened by dredging
to elevation 13 m to pro-vide additional storage capacity. After
the completion of embankmentconstruction, sod was placed over the
embankment (Figure 10c).
Six SPT conrmation borings were performed on August 16 and
17,2007 along the Cell 7 South embankment crest as a part of
thepost-construction conrmation process. The borings generally
encoun-tered layers of compacted sand and variably silty sand.
However, pocketsof tailings were encountered between Elevations
+2.3 and +4 m, 1and +1.5 m, and 2 to 3 m during the performance of
one of theseborings. Recorded SPT blow counts in the tailings
ranged from 2 to 13blows per foot, indicating a very soft to stiff
consistency for the soil unit.This worst case scenario was modeled
at the permitted drawdown rate
uilt worst case scenario. (b) The critical failure plane.
-
2007
2007
2008
2008
2008
2009
2009
te
e re
98 A.T. zer, L.G. Bromwell / Engineering Geology 151 (2012)
8999and the calculated factor of safety against rapid drawdown was
slightlyreduced from 2.8 (Figure 9) to 2.5 (Figure 11) which still
satised theminimum required FOS against rapid drawdown (1.3).
Cell 7 embankment construction was completed on August 2007.As
shown in Fig. 12, the reservoir has been through 4 drawdowncycles
since completion. The average drawdown rate was approxi-mately 12
cm/day which was less than the design drawdown rateof 30 cm/day.
Maximum daily drawdown during the last 4 yearswas 25 cm/day which
is relatively close to the design drawdownrate. Frequent visual
inspection of the physical condition of thesouthern embankment of
Cell 7 has been performed since the initiallling and during
subsequent drawdowns. No indications of struc-tural instability
have occurred indicating that the toe buttress and
07/0
6/
05/1
0/
02/0
2/
01/0
6/
29/0
9/
27/0
1/
27/0
5/
Fig. 12. Post-construction performanc-1
0
1
2
3
4
Res
ervo
ir Po
ol E
leva
tion
(m, N
AVD)
Daremoval of the silt/clay tailings materials 6 to 9 m inside
the toehas provided structural integrity. In addition to the visual
observa-tions, post construction behavior of the embankment is
alsomodeledunder the historical pool levels (Figure 12). Post
construction analy-ses and observations indicate that themeasures
taken to increase thestability against rapid drawdown loading were
successful.
8. Conclusions
This paper describes the design of Cell 7 of the L8SR in
westernPalm Beach County, Florida, part of which was built on
recentlysedimented tailings consisting of low strength, low
plasticity silt/clay, and variably silty sand. Fully coupled steady
state and transientseepage models with slope stability analyses
were performed for op-erational and rapid drawdown conditions,
using both LEM and FESBMtechniques. The results of the analyses
showed that a LEM analysisusing a method that satises both force
and moment equilibrium(i.e., Spencer's Method) agreed closely with
more complex FESBMbased models.
Additionally, results using a linear elastic soil behavior model
forFESBM analysis were compared with more complex
constitutivemodels for the low plasticity silt/clay and variably
silty sand tailings.The linear elastic soil behavior model gave
comparable and reason-able values of stress distribution during
both steady state and rapiddrawdown analyses. However, for the
modeling of higher plasticitytailings materials, more complex
material models should be studied.Excess pore pressure response
during rapid drawdown was com-pared between transient seepage and
stress/pore pressure models,and it was found that both approaches
produced nearly identical re-sults. Consequently, the coupled
stress/pore pressure model usinglinear-elastic material properties
resulted in reasonable values ofstress distribution during rapid
drawdown. It was found that rapiddrawdown governs the design, and
regardless of the type of stabilityanalysis (either LEM or FESBM),
undrained DSS strength test resultsfor the low plasticity tailings
should be considered for the designagainst rapid drawdown failure.
It would be prudent for a design en-gineer to perform the analyses
using the results from both triaxial andDSS tests to dene the shear
strength behavior of undrained soils, anduse the lower, more
critical, factor of safety value for the design.2.4
2.45
2.5
2.55
2.6
2.65
2.7
2.75
2.8
2.85
2.9
24/0
9/20
09
22/0
1/20
10
22/0
5/20
10
19/0
9/20
10
17/0
1/20
11
17/0
5/20
11
14/0
9/20
11
Fact
or o
f Saf
ety
Reservoir elevationFactor of Safety
garding factor of safety calculations.References
Alkasawneh, W., Malkavwi, A.I.H., Nusairat, J.H., Albataineh,
N.A., 2008. Comparativestudy of various commercially available
programs in slope stability analysis. Com-puters and Geotechnics
35, 428435.
Aubertin, M., Mbonimpa, M., Bussiere, B., Chapuis, R.P., 2003. A
model to predict thewater retention curve from basic geotechnical
properties. Canadian GeotechnicalJournal 40, 11041122.
BCI Engineers and Scientists, 2007. Design Basis Memorandum: L8
Storage ReservoirPalm Beach County, FL. Prepared for Palm Beach
Aggregates, Inc.
Belirgen, M.M., 2007. Investigation of stability of slopes under
drawdown conditions.Computers and Geotechnics 34, 8191.
Britto, A.M., Gunn, M.J., 1987. Critical state soil mechanics
via nite elements. JohnWiley and Sons, Inc., New York.
Bromwell, L.G., Ying, K., 2009. Predicted vs. measured slope
erosion at L-8 reservoirduring hurricane Jeanne. Association of
State Dam Safety Ofcials (ASDSO) DamSafety'09, Hollywood, FL.
Bromwell, L.G., Ying, K., 2010. Correlation of predicted and
measured slope erosion.ISCE-5, 5th International Conference on
Scour and Erosion, San Francisco, CA.
Christmann, C.W., Bromwell, L.G., Schwartz, M., 2009.
Soil-bentonite slurry wall forseepage control at L-8 reservoir.
Association of State Dam Safety Ofcials(ASDSO) Dam Safety'09,
Hollywood, FL.
Dawson, E.M., Roth, W.H., Drescher, A., 1999. Slope stability
analysis by strength reduc-tion. Geotechnique 49 (6), 835840.
Duncan, J.M., 1996. State of art: limit equilibrium and nite
element analysis of slopes.Journal of the Geotechnical Engineering
Division, ASCE 122 (7), 577596.
Duncan, J.M., Chang, C.Y., 1970. Nonlinear analysis of stress
and strain in soils. Journalof the Soil Mechanics and Foundation
Engineering Division, ASCE 96 (SM5),16291653.
Duncan, J.M., Wright, S.G., Wong, K.S., 1990. Slope stability
during drawdown. (May1990) Proceedings of the H. Bolton Seed
Memorial Symposium, vol. 2, pp. 253272.
Fredlund, D.G., Xing, A., 1994. Equations for the soilwater
characteristics curve. Cana-dian Geotechnical Journal 31,
521532.
-
Geo-Slope International Ltd., 2008a. Seepage Modeling with
SEEP/W 2007, An Engi-neering Methodology, 3rd edition. Geo-Slope
International Ltd., Calgary: Alberta,Canada.
Geo-Slope International Ltd., 2008b. Stability Modeling with
SLOPE/W 2007, An Engi-neering Methodology, 4th edition. Geo-Slope
International Ltd., Calgary: Alberta,Canada.
Geo-Slope International Ltd., 2008c. Stress-deformation Modeling
with SIGMA/W2007, An Engineering Methodology, 3rd edition.
(Calgary, Alberta, Canada).
Grifts, D.V., Lane, P.A., 1999. Slope stability analysis by nite
elements. Geotechnique49 (3), 387403.
Huang, M., Jia, C.-Q., 2009. Strength reduction FEM in stability
analysis of soil slopessubjected to transient unsaturated seepage.
Computers and Geotechnics 36, 93101.
Jiang, Y.S., 1990. Slope Analysis Using Boundary Elements.
Springer-Verlag Publishers,New York.
Krahn, J., 2003. The 2001 R.M. Hardy Lecture: the limits of
limit equilibrium analyses.Canadian Geotechnical Journal 40,
643660.
Krahn, J., 2007. Limit equilibrium, strength summation and
strength reductionmethods for assessing stability. Proceedings of
1st Canada-U.S. Rock MechanicsSymposium, Vancouver, B.C., Canada,
pp. 2830.
Ladd, C.C., DeGroot, D.J., 2003. Recommended practice for soft
ground site characterization:Arthur Casagrande Lecture. 12th
Panamerican Conference on Soil Mechanics and Geo-technical
Engineering, MIT, June 2225.
Lambe, T.W., Silva, F., 2003. Evaluating the stability of an
earth structure. Proceedingsof 12th Pan-American Conference on Soil
Mechanics and Foundation Engineering,Cambridge, MA, pp.
27852790.
Land, L.F., Rodis, H.G., Schneider, J.J., 1973. Appraisal of the
Water Resources of EasternPalm Beach County Florida. U.S. Geologic
Survey.
Londono, A.C., 2004. Bank instability resulting from rapid ood
recession along theLicking River, Kentucky. Master of Science
Thesis, University of Cincinnati.
Lowe, J., Karaath, L., 1960. Stability of earth dams upon
drawdown. Proceedings of theFirst PanAmerican Conference on Soil
Mechanics and Foundation Engineering,Mexico D.F, pp. 537552.
Matsui, T., San, K., 1992. Finite element slope stability
analyses by shear strength re-duction technique. Soils and
Foundations 32, 5970.
Morgenstern, N.R., Price, V.E., 1967. The analysis of the
stability of embankments as-suming parallel interslice forces.
Geotechnique 17, 1126.
zer, A.T., Faulk, W., Bromwell, L.G., 2009. Rapid drawdown
stability of dam on silt/claytailings foundation. Association of
State Dam Safety Ofcials (ASDSO) Dam Safe-ty'09, Hollywood,
Florida, USA.
Petuch, E.J., Roberts, C.E., 2007. The Geology of the Everglades
and Adjacent Areas. CRCPress, Boca Raton, Florida.
Schanz, T., Vermeer, P.A., Bonnier, P.G., 1999. The hardening
soil model: formulationand verication. Beyond 2000 in Computational
Geotechnics 10 Years of PLAXIS.Balkema, pp. 281296.
Spencer, E., 1967. A method of analysis of embankments assuming
parallel inter sliceforces. Geotechnique 17 (1), 1126.
Unites States Army Corps of Engineers (USACE) Engineering and
Design, 2003. SlopeStability Manual, No. EM 1110-2-1902.
Van Genuchten, M., 1980. A closed-form equation for predicting
the hydraulic conduc-tivity of unsaturated soils. Soil Science
Society of America Journal 44, 892898.
Wright, S.G., 1969. A study of slope stability and the undrained
shear strength of clayshales. Ph.D. Dissertation, University of
California, Berkley, CA.
Wright, S.G., Duncan, J.M., 1987. An examination of slope
stability computation proce-dures for sudden drawdown.
Miscellaneous Paper GL-87-25. U.S. Army EngineerWaterways
Experiment Station, Vicksburg, MS.
Yang, H., Rahardjo, H., Leong, E.-C., Fredlund, D.G., 2004.
Factors affecting drying andwetting soilwater characteristic curves
of sandy soils. Canadian GeotechnicalJournal 41, 908920.
Zandarin, M.T., Oldecop, L.A., Rodriguez, R., Zabala, F., 2009.
The role of capillary waterin the stability of tailing dams.
Engineering Geology 105, 108118.
99A.T. zer, L.G. Bromwell / Engineering Geology 151 (2012)
8999
Stability assessment of an earth dam on silt/clay tailings
foundation: A case study1. Introduction2. Site geology3. Field
exploration program4. In-situ and laboratory testing5. Seepage and
stability analyses methods used5.1. Seepage analyses5.2. Stability
analysis
6. Seepage and stability analyses results6.1. Steady state
seepage and long-term stability analysis results6.2. Rapid drawdown
stability analysis results
7. Construction details and post-construction performance8.
ConclusionsReferences