Assessment of Slope Stability: Serge - Leroueil@gci - Ulaval.ca

Downloaded from ascelibrary.org by dml u on 03/08/14. Copyright ASCE. For personal use only; all rights reserved.
Assessment of Slope Stability

Serge Leroueil1 and Luciano Picarelli2
1
Department of Civil and Water Engineering, Laval University, Quebec City, PQ,
Canada, Serge.Leroueil@gci.ulaval.ca
2
Department of Civil Engineering, Secunda Universit di Napoli, Naples, Italy
ABSTRACT: Assessing the stability of slopes is a difficult task that requires a
rigorous approach. This paper starts by presenting a geotechnical characterization of
slope movements that helps in classifying the information related to slopes. The paper
then shows the tremendous progresses that have recently been made for monitoring
slopes, understanding slope behavior, and for numerically simulating coupled hydromechanical problems in saturated/unsaturated soils. A methodology for assessing the
stability of slopes is presented. It includes qualitative and quantitative approaches and,
separately, the characterization of the post-failure stage. Finally, there is some
discussion on the use of factor of safety, use of numerical models and consideration of
risk.
INTRODUCTION
Assessing the stability of existing slopes is probably the most difficult task of the
geotechnical engineer. Literature generally refers to landslides that have occurred but
engineers reality is most of the time to assess the stability of existing, still standing
slopes. Stating on the stability of a given slope is also an important responsibility for
the engineer as excessive movements may have important safety and economic
consequences.
It is clear that the first and essential condition for assessing the stability of a slope is
the understanding of the mechanical processes that lead or may lead to movements or
failure. After general remarks on the organization of the information on slopes, recent
developments, in particular on methods of survey and monitoring, understanding of
slope behavior and numerical modeling, are described. Finally the paper proposes a
methodology for the assessment of slope stability in practice. A priori, the approach
presented concerns natural slopes, but could be adapted to engineered slopes. Also,
even if some aspects are of general use, this paper concerns mostly slopes in soils.
Due to the limited space allowed, this paper is not a State-of-the-Art on the
122
Page 1
Geotechnical Engineering State of the Art and Practice
GEOTECHNICAL ENGINEERING STATE OF THE ART AND PRACTICE 123
assessment of slope stability, but rather a set of remarks that seem important to the
authors. For the same reason, the number of citations has been limited and the reader
is encouraged to examine them to find additional references. It is worth mentioning
that in a recent book, Cornforth (2005) covers many aspects of slope engineering and
that other books on landslides are in various stages of preparation.
ORGANIZATION OF THE INFORMATION ON SLOPES
To help in better understanding and analyzing slope movements, and organizing the
knowledge on slope behavior, Vaunat et al. (1994) and Leroueil et al. (1996) proposed
a Geotechnical characterization of slope movements that can be schematized as a 3D matrix (Fig. 1) with the three axes corresponding to: types of movement, types of
material involved, and stages of movement. The types of movement are those
proposed by Cruden et al. (1994) and Cruden and Varnes (1996), i.e. falls, topples,
slides, lateral spreads and flows. For materials, Varnes (1978) considered 3 main
classes, i.e. rock, debris and earth. They are not sufficient to describe and take into
account the mechanical behavior of the geomaterials, and more detailed classifications
have been proposed (e.g. Fig. 2). Even such classification is not sufficient and it may
be important to specify if the sand is loose or dense (contractant or dilatant),if the soil
is saturated or unsaturated, etc. It is also suggested dividing slope movements in four
stages that are associated with their own controlling laws and parameters (Fig. 3): the
pre-failure stage, including all the deformation processes that may occur before
failure, even if this latter never happens; the onset of failure, characterized by the
formation of a continuous shear surface (slip surface) through the entire soil mass as
this is the fundamental mechanism of failure in soils (it is different in the case of rock
masses); the post-failure stage, which includes movement of the soil mass involved in
the landslide, from just after failure until it essentially stops (flows are thus postfailure movements); the reactivation (or active) stage, when a soil mass slides along
one or several pre-existing shear surfaces. At an active or reactivated stage, the
residual strength is mobilized along the slip surface, which is different than for a firsttime failure. The International Committee on Landslides (JTC1) is presently working
on a revision of the classification proposed by Varnes (1978) and Cruden and Varnes
(1996).
Fig. 1. Schematic slope movement characterization (from Leroueil et al., 1996)
Page 2
124 GEOTECHNICAL ENGINEERING STATE OF THE ART AND PRACTICE
Fig. 2. Material types considered in the slope movement characterization (from

Leroueil, 2001)
Fig. 3. Different stages of slope movements (from Leroueil et al., 1996)

For each relevant element of the characterization matrix, the information can be put
in a characterization sheet (Fig. 1) with:
- The laws and parameters controlling the phenomenon; it could be the strength
parameters at the failure and reactivation stages.
- The predisposition factors (in mechanical terms, they are essentially the initial
conditions) that give information about the present situation and determine the
slope response following the occurrence of a triggering factor (e.g. presence of
a loose sand layer in a seismic area).
- The triggering factors that lead to failure, or aggravating factors that produce a
significant modification of stability conditions or of the rate of movement.
They can be temporary (e.g. heavy rainfall or rapid drawdown) or progressive
(e.g. erosion or weathering), (see Table 1).
- The revealing factors that provide evidence of slope movement but generally
do not participate to the process (e.g. presence of cracks).
- The possible, direct or indirect, consequences of the movement.
Page 3
Table 1. Common triggering or aggravating factors (Leroueil, 2004)

Terrestrial
Submarine
x
x
x
x
x
x
x
x
x
x
x
Increase in shear stress

Erosion and excavation at the toe
Surcharging at the crest, sedimentation (and possible underconsolidation)
Rapid drawdown of water level adjacent to slopes
Earthquake
Volcanic activity
Fall of rock
Decrease in strength
Infiltration due to rainfall, snow melt, irrigation, water leakage
from utilities
Weathering
Physico-chemical changes
Gas hydrate dissociation
Pile driving
Fatigue due to cyclic loading and creep
Thawing of frozen soils
Possible increase in shear stress and decrease in strength
Vibrations and earthquake shaking that can generate excess pore
pressures
Swinging of trees due to wind gusts
Storm waves and changes in sea level
x
x
x
x
x
x
x
x
x
x
x
x
Another concept that helps in analyzing engineering problems and that is

increasingly used is the concept of risk. Important works have been performed in
recent years for its applicability to slope engineering, and specialty conferences on
Landslide Risk Assessment and Landslide Risk Management were respectively
held in Honolulu in 1997 and in Vancouver in 2005.
The total risk, RT, is defined by Varnes et al. (1984) as the set of damages resulting
from the occurrence of a phenomenon. It can be described as follows:
RT =
6HRiVi
[1]
in which H is the hazard or the phenomenon occurrence probability; Ri (for i = 1 to n)

are the elements at risk (persons, buildings, infrastructures, etc.) potentially damaged
by the phenomenon; and Vi is the vulnerability of each element Ri, represented by a
damage degree between 0 (no loss) and 1.0 (total loss). In the context of the
geotechnical characterization of slope movements (Fig. 1), the elements at risk and
their vulnerability should be, directly or indirectly, in Movement consequences
whereas the hazard, probability that the triggering factor may reach a given value that
would cause a failure, should be in Triggering or aggravating factors. Also, in the
context of slopes, the hazard has to be subdivided into two parts: the hazard associated
with the possibility of having a failure, Hf; and the hazard associated with the
possibility that the post-failure stage presents specific characteristics (e.g. debris
reaching a given distance), Hpost-f; H = Hf x Hpost-f.
Page 4
RECENT DEVELOPMENTS
Important recent developments can be divided in three classes: technologies for
surveying and monitoring slopes; achievements about the understanding of slope
behavior; and the development of procedures for numerical modeling of slope
behavior. The technologies for improving the safety of slopes could be added, but are
not considered here.
Methods of survey and monitoring
Table 2 lists the main instruments and technologies that can be used for surveying and
monitoring slopes.
Table 2. Main instruments and technologies for surveying and monitoring slopes
Displacements and deformations
Crackmeters; extensometers
Tiltmeters
Inclinometers
SAA (shape-accel-array)
Photogrammetry
GPS and dGPS
Laser ranging or Electron. Dist. Meas.
(EDM)
InSAR (Interferometric Synth. Apert.
Radar)
Optic fibers
Acoustic emission/micro seismic monitoring
system (AE/MS)
Topography
Aerial photogrammetry
LIDAR
Multibeam echosounding (marine environ.)
Environmental conditions
Weather station
Thermistors
Pore water content and pressure

TDR (Time Domain Reflectometry)
Piezometers
Tensiometers
Psychrometers
Some remarks can be made in relation with new systems:

- Very important developments have occurred in recent years in terms of
technologies, possibilities to take, send and store readings, and interpretation
methods. Most of the instruments can now be monitored automatically, almost
continuously if necessary, and the data can be sent via wireless links to an
interpretation and analysis center. Some of these developments are specified
below.
- Most of the instruments or systems listed in Table 2 have been used on one of
the sites indicated below and the reader can refer to the references for details.
- The information automatically obtained (rainfall intensity, pore pressures, rate
of displacement, etc.) can be used for landslide triggering prediction and in
early warning systems (Picarelli et al., 2009).
- Inclinometers: The most common procedure is to install an inclinometer casing
and take readings at different times with a probe. It is however possible to
leave one or several inclinometer segments in-place at selected depths and take
regular readings.
- SAA (for Shape-Accelerometer-Array; see Rollins et al., 2009, and Cloutier et
Page 5
al., 2011): It is a rope-like array of sensors and microprocessors that fits in a

small casing. It is made of small segments that are connected by flexible joints
and contain three accelerometers. SAA measures deformations (and also
vibrations) and can be used vertically (as an in-place inclinometer) or
horizontally (Cloutier et al., 2011).
GPS and dGPS (for Differential Global Positioning System): The accuracy of
dGPS is of a few millimeters.
EDM (for Electronic Distance Measurement) or Laser ranging: The system
uses laser beams and mirrors from a large distance to the target (possibly a few
kilometers) and, according to the web site of Turtle Mountain
(www.ags.gov.ab.ca/geohazards/turtle_mountain), the accuracy would be of
one part per million, i.e. one millimeter over a distance of one kilometer.
InSAR (for Interferometric Synthetic Aperture Radar) and DInSAR: This
technology uses radar waves and can be ground-based or satellite-based
(Singhroy and Molch, 2004; Couture et al., 2011). In the case of DInSAR, the
accuracy is of a few millimeters. Interestingly, it is possible to obtain images
back to the early 90s, and thus to establish movement history for the last 20
years or so (T.R.E., www.treuropa.com).
Optic fibers: Originally used to measure strains in structural elements, the use
of optical fibers for monitoring strains in geotechnical engineering is spreading
(Shi et al., 2008; Picarelli and Zeni, 2009). Monitoring is performed by
stimulated Brillouin scattering (SBS) which enables to obtain temperature and
strain variations along a single-mode optical fiber through Brillouin shift
measurements (Olivares et al., 2009). The advantage of optical fibers is
twofold: the possibility to measure changes in temperature or strains
everywhere along a fiber, even over long distances, and their extremely low
cost.
AE/SM (for Acoustic Emission and Micro Seismic Monitoring System): Used
in rocks and soils to detect AE generated by inter-particle friction, fracture
propagation and displacements along discontinuities (Dixon et al., 2003; Dixon
and Spriggs, 2007; Eberhardt et al., 2004; Amitrano et al., 2010). However, as
in most soils AE level is low and attenuation is high, Dixon et al. (2003)
proposed to use waveguides (tubes in a preformed borehole with noisy
material (sand or gravel) around the tube) in order to amplify the noises. Dixon
and Spriggs (2007) show that the AE signal is linked to the deformation rate.
LiDAR (for Light Detection and Ranging): Can be airborne or on the ground,
and can be used on large areas. It penetrates vegetation and can provide a highresolution topography of the ground surface that can be used to generate a 3-D
terrain model. The accuracy is in the order of 0.15 m.
Weather station: It may include the measurement of temperature, relative
humidity, rainfall, wind speed and solar radiation. These parameters may allow
the determination of evapo-transpiration potential.
TDR (for Time Domain Reflectometry): It is generally used for determining
the average water content around a probe. The experimental device consists of
an electromagnetic pulse generator connected, through a coaxial cable, to a
metallic probe a few decimeters long, which is buried in the soil. An
Page 6
electromagnetic pulse is sent through the soil, and the reflected signal is
acquired, providing the soil volumetric water content. Greco (2006) developed
an interesting inverse procedure to retrieve the water content profile along the
probe and not only an average value. Figure 4 shows an example from
infiltration tests carried out on a model slope subjected to artificial infiltration
until failure (Greco et al., 2010).
Piezometers (for positive pore pressure measurements): The selection of the
type of piezometer should be decided mostly on the basis of their time lag and
the monitoring method (manual or remote).
Tensiometers: Tensiometers and other probes such as psychrometers are used
to monitor negative pore pressures in saturated and unsaturated soils, in
particular granular soils subjected to precipitations. Associated with water
content measurements, tensiometer readings can be used to define the in situ
Water Retention Curve of the soil (Comegna et al., 2011).
10
t = 0 min
t = 6 min
t = 12 min
t = 15 min
6
z
[cm]
t = 19 min
t = 21 min
t = 23 min
2
0
0.20
t = 26 min
t = 29 min
0.30
0.40
0.50
[m3/m3]
0.60
0.70
0.80
Fig. 4. Volumetric water content profiles obtained from TDR readings (modified
after Greco et al., 2010).
The examples given below illustrate the use that can be made of monitoring systems.
Cloutier et al. (2011) and Couture et al. (2011) present instrumentation and
monitoring data at the Gascons rockslide, Quebec. The monitoring system includes
crackmeters, extensometers, tiltmeters, piezometers, inclinometers and SAA system
installed horizontally, with most of these instruments remotely monitored. Surface
displacements are monitored by satellite-based InSAR technology, using a special
technique (PTA-InSAR technique) to improve the accuracy. Figure 5 shows the site
with examples of monitoring data. Other examples of very well instrumented sites in
rock are the Turtle Mountain Monitoring Project and Field Laboratory in Alberta,
Canada, (Froese and Moreno, 2007), the Randa site, Switzerland, and the Aknes site,
Norway (see Eberhardt et al., 2008).
Page 7
Fig. 5. The Gascons rockslide with displacement profiles obtained from SAA,
pore pressures versus time, and displacements of corner reflectors obtained by
InSAR technique versus time (modified after Cloutier et al. (2011) and Couture
et al. (2011)).
Since 2002 a steep calcareous slope mantled by unsaturated pyroclastic soils subjected
to a rapid killer flow-slide (1999) at Cervinara, Italy, has been monitored to investigate
the slope-atmosphere interaction (effects of precipitations, and of air temperature and
humidity on infiltration and evapo-transpiration, thus on water content and suction
values in the soil). The monitoring system includes a weather station, a number of
tensiometers and a number of TDRs, with some couples of TDRs and tensiometers at
the same locations and depths in order to correlate values of the volumetric water
content and suction. Figure 6 shows the layout of a part of the instrumentation, which
is powered by solar panels for automatic data acquisition every 60 minutes, and some
results (Comegna et al., 2011). The instrumentation allows the estimate at any time of
the safety factor of the slope.
Understanding of slope behavior
Understanding the mechanisms of slope movement is essential for evaluating the
stability conditions of slopes, the post-failure stage, the possible consequences of a
failure, and selecting mitigation methods. So, these processes need to be investigated.
The following outcome can be noted.
Infiltrations and evapo-transpiration. The pore water pressures and the physical state
of soil (water content, degree of saturation, etc.) generally vary with alternating rainy
and dry periods. The processes involved are complex but progresses have recently
been made due to a better understanding of the hydraulic and mechanical behavior of
Page 8
unsaturated soils, and also to the development of numerical hydraulic and hydromechanical models. The analysis of infiltration in an unsaturated soil requires the soil
water retention curve (relationship between degree of saturation and matric suction),
the hydraulic conductivity versus suction relationship and, if the geomaterial is not
assumed perfectly rigid, a compressibility law. Collins and Znidarcic (2004) examined
the case of infiltration in an infinite slope consisting of 4 m of an unsaturated soil
deposit over a drained boundary with zero suction. Figure 7 shows the water pressure
head profiles at different times for a fine grain soil and a coarse grain soil
characterized by typical permeability functions. Due to higher hydraulic conductivity
in this latter case, infiltration progresses more rapidly. However, in both cases, it takes
a significant amount of time before the water front reaches the bottom of the soil layer.
Collins and Znidarcic (2004) also examined the stability of slopes, considering
suction-dependent shear strength, in particular for defining the critical depth. Similar
studies have been performed by other researchers (e.g. Ng and Shi, 1998; Cai and
Ugai, 2004; Lu and Godt, 2008).
Fig. 6. Measurement of rainfall (a and e), suction (Tensiometers; c and g) and

volumetric water content (TDRs; b and f) at Cervinara (modified after Comegna
et al., 2011).
It is also worth noting that several natural slopes have been recently instrumented
and monitored (Springman et al., 2003; Ochiai et al., 2004; Rahardjo et al., 2005; Papa
et al., 2009; Damiano et al., 2012). These studies give useful information regarding the
infiltration process, the influence of antecedent rainfall, the influence of
heterogeneities, the percentage of rainfall contributing to infiltration, the evapotranspiration, etc., providing interesting data sets for testing numerical models.
Pore pressure changes in soil deposits. Even in saturated conditions, because of the
compressibility of soils, a change in pore water pressure at the boundaries is not
Page 9
reflected instantaneously to the entire soil deposit. Changes require a

consolidation/swelling process, controlled by the coefficient of consolidation/swelling
of the soil. This can be illustrated by observations made at Wabi Creek, Ontario, in a
soft clay deposit (Kenney and Lau, 1984; Fig. 8). At some distance from the slope the
annual variation of the water table level is in excess of 2 m, but the variation in pore
pressure at depths larger than 10 m or so is less than 0.5 m; it corresponds to a
coefficient of consolidation/swelling in the order of 10-6 m2/s. Similar observations
were made by Demers et al. (1999). According to Vaughan (1994), the seasonal
fluctuations in pore pressure in stiff clays from UK only extend down to about 4 m.
So, in most clayey deposits, pore pressures are continuously varying and are not
hydrostatic (Leroueil, 2001). This also suggests that weather-induced landslides are
generally shallow, and deep failures can be due to different causes or complex hydromechanical processes.
Fig. 7. Effect of vertical water infiltration in a column of fine grain soil (k sat = 1.5
x 10-8 m/s) (a) and coarse grain soil (ksat = 1.5 x 10-6 m/s) (b) (modified after
Collins and Znidarcic, 2004).
Slope instability due to pore pressure increase. When a loose soil specimen of
granular soil is subjected to an undrained loading, it reaches a peak on a line in a stress
diagram (IL, between D and C on Fig. 9a) called Instability Line by Lade (1993), and
then moves towards its critical state (C on Fig. 9a), on the critical state line (CSL)
characterized by Ics.
When a slope is subjected to pore pressure increase due to infiltration or rising water
table, total stresses and shear stresses remain almost constant but effective stresses
decrease. In a stress diagram such as the one in Fig. 9a, this corresponds to a stress
path such as from Ils towards Yls. If the stress path reaches the instability line at Yls,
the soil has a tendency to move towards its critical state C. As the deviatoric stress at
C is smaller than that due to gravity forces in the slope (q at Ils), there will be static
liquefaction of the soil and collapse of the slope. Major consequences of this
phenomenon are that failure is triggered at an angle of strength mobilization smaller
than Ics, and that instability (at Yls) is followed by an increase in pore water pressure
since p decreases, with a significant part of the potential energy available at the onset
Page 10
of failure released into kinetic energy.

Chu et al. (2003) showed that the concept of instability (development of plastic
deformations) applied to slopes can be extended to dense soils, with the instability line
extending above the CSL (C towards Yds on Fig. 9a) and corresponding to a dilatant
soil behavior. As shown by Wang and Sassa (2001), Damiano (2003) and more
generally by Leroueil et al. (2009), flume tests and field evidences confirm this model
for both loose and dense soils.
Fig. 8. Location of piezometers and annual changes of piezometric level at Wabi

Creek (from Kenney and Lau, 1984).
Fig. 9. (a) Stress path in sand followed at q = cst due to pore pressure increase
(from Leroueil, 2004) and (b) Comparison between stress path in slope subjected
to pore pressure increase and that implicitly assumed in stability analysis.
Page 11
Progressive failure. If there is: (a) a brittle soil, i.e. a soil presenting a peak and a
strain-softening behavior either in drained or undrained conditions; (b) non-uniformity
in the distribution of shear stresses; (c) shear stresses that locally reach the peak
strength of the soil; and (d) boundary conditions such that strains may develop,
progressive failure will develop (see Urciuoli et al., 2007). As such conditions exist in
many soils and slopes, progressive failure is quite common. Potts et al. (1997) studied
the case of cut slopes in London clay considering Ip = 20 and cp = 7 kPa for the
peak strength and Ir = 13 and cr = 2 kPa for residual conditions. These authors
explained delayed failures in cut slopes and concluded that the average strength
parameters at failure would be a friction angle of 18 and zero cohesion, which are
parameters close to those back-calculated from first-time failures (I = 20 and c = 1
kPa; Chandler, 1984). These average mobilized parameters are intermediate
between peak and large deformation parameters which is typical of progressive
failure.
Bernander (2000) indicated that a number of large landslides that occurred in
Sweden were associated with a perturbation upslope and progressive failure, from
upslope towards downslope. In eastern Canada on the other hand, local failure is
generally initiated at the toe of the slope, generally by erosion; a failure surface then
progresses horizontally over large distances into the soil mass and the soil above the
failure surface adjusts to these new conditions, generally by forming horsts and
grabens, typical of spreads. Limit equilibrium stability analysis methods performed
either in drained or undrained analyses provide factors of safety in these spread
landslides that are much higher than 1.0 and thus cannot be used. In fact, spreads are
not well understood yet and are the object of researches by studying case histories in
detail and performing numerical simulations (Locat et al., 2011; Locat, 2012).
It has also been shown that progressive failure may develop and explain failures in
rock slopes (Eberhardt et al., 2004) and also in earth-dams, at Carsington, UK, (Potts
et al., 1990) and at Aznalcollar, Spain, (Gens and Alonso, 2006). It is thus quite a
common process, however generally ignored in practice.
Earthquake induced landslides.
The effects of earthquakes on the stability of slopes depend mostly on their magnitude,
duration and frequency content, and on the main structural and mechanical
characteristics of the rock/soil mass. Comprehensive reports on the effects of
earthquakes have been extremely useful to develop our knowledge of the problem
(e.g. Berz et al., 1980; DElia et al., 1985; Towhata et al., 2001; Ishihara, 2005; Lin et
al., 2008; Wang et al., 2009; Locat, 2011). These studies have highlighted significant
aspects, as the delay sometimes observed between the earthquake occurrence and the
onset of slope failure (Seed et al., 1973; DElia et al., 1985; Seid-Karbasi and Byrne,
2007), the role of cyclic soil liquefaction in the triggering of catastrophic landslides,
especially in saturated sands (e.g. Ishihara, 1993; Idriss and Boulanger, 2008)) and
some complex post-cyclic lateral spreads in saturated clay (Fenelli et al., 1992;
Olivares, 1997; Idriss and Boulanger, 2008). Also, performing two-dimensional
dynamic response analyses, Ashford and Sitar (1994) showed that the peak
acceleration computed at the crest of the slope is larger than the free-field value.
Finally, field experience has been extensively used for defining conditions that can
Page 12
trigger liquefaction (Youd et al., 2001) as well as the residual undrained shear strength
of liquefied sands (Seed and Harder, 1990; Olson and Stark, 2002).
Studies have been performed about the effect on the stability of a slope of a low
permeability soil layer in or above a liquefiable layer of sand. In case of liquefaction
by shaking, the lower part of the liquefied sand layer densifies whereas the upper part,
thus beneath the low permeability layer that acts as a barrier, expands. An extreme
case of expansion is the formation of a water film immediately below the low
permeability layer. As a result of this expansion, the shear strength decreases (to zero
in the case of a water film) and there may be shear localization beneath the low
permeability layer and slope failure. This has been observed experimentally in shaking
tests (Kokusho, 1999) and in centrifuge tests (Kulasingam et al., 2004; Phillips and
Coulter, 2005; Malvick et al., 2008), and numerically simulated (Seid-Karbasi and
Byrne, 2007). Figure 10 shows a slope model of loose sand with a silt layer seen in
lighter grey, in a centrifuge that was subjected to shaking (Malvick et al., 2008).
Whereas the localization is clearly observed immediately beneath the silt layer about
120 s after shaking (Fig. 10c), it has not developed during shaking (Fig. 10b). This is
explained by the fact that the redistribution of voids, and thus the change in strength,
requires time. Because many natural soil deposits are stratified, it is thought that this
process is the explanation for the failure of numerous submerged slopes, in particular
for the failure of gentle slopes of a few percent. Details on the process and
implications are given by the previously mentioned authors, in particular Kokusho
(2003) and Malvick et al. (2008); it is emphasized in particular that conventional
laboratory tests on undisturbed soil samples cannot reproduce this phenomenon and
provide strength that cannot be representative of the strength of the soil beneath the
low permeability layer.
Fig. 10 Photographs of centrifuge model of Nevada sand with embedded silt arc
(lighter layer) subjected to earthquake motion (modified from Malvick et al.,
2008).
Page 13
Numerical modeling
Numerical models for hydraulic or coupled hydro-mechanical problems have been
developed for both saturated and unsaturated soil conditions; they constitute very
powerful tools for studying a variety of problems and analyzing specific, sometimes
complex case studies. Applications can be found in a variety of domains: infiltration in
saturated and unsaturated conditions (e.g. Rahardjo et al., 2008); excavations in clayey
deposits (e.g. Kovacevic et al., 2007); rapid drawdown (Pinyol et al., 2008 and 2011);
slope subjected to earthquake (e.g. Byrne et al., 2006; Takahashi et al., 2008); flowlike landslides (e.g. McDougall and Hungr, 2004; McDougall et al., 2008). These
developments have given a more detailed insight to a number of problems that were
only coarsely understood. This can be illustrated by some examples.
Infiltration. The development of advanced theories and models about the behavior of
unsaturated soils have favored the development of modern codes, working at slope and
regional scales, for the analysis of the soil response to infiltration. SEEP (Geoslope,
2004) and CODEBRIGHT (Saaltink, 2005) are examples of codes working at slope
scale that can relate the stability of slopes to precipitation via infiltration analysis.
Codes conceived to perform analyses at regional scale start from a terrain model (e.g.
TRIGRS by Baum et al., 2008). Even though losing the possibility to perform
sophisticated infiltration analyses, they give the possibility to setting up reliable early
warning systems in areas subjected to risk of rainfall-induced landslides and to
produce in real time dynamic risk mapping by coupling infiltration analysis with
stability analysis. Along this way, researchers of the Seconda Universit di Napoli
developed the so called I-MODE 3D Finite Volume Code (Olivares and Tommasi,
2008). It is used as a basic component of a modeling chain which includes the
COSMO-LM code for weather forecasting. The geotechnical I-MODE 3D calculates
the effects of precipitation, i.e. the increases in water content and decreases in suction
or positive pore pressure, and performs stability analyses under the hypothesis of an
infinite slope in unsaturated soils, using an extension of the Mohr-Coulomb failure
criterion to unsaturated soils (Fredlund and Rahardjo, 1993). For the analysis of the
hydraulic effects, the Water Retention Curve of the soil based on the Van Genuchten
expression (1980) and the hydraulic conductivity based on the Brooks and Corey
expression (1964) are used. Some applications are reported by Olivares and Tommasi
(2008) and by Damiano and Olivares (2010).
Temporary cut in clay. Kovacevic et al. (2007) considered the process of pore pressure
equilibration and progressive failure for minimizing the volume of an open-cut
excavation in London clay for the below-ground construction of the London Heathrow
Airports Terminal 5. This excavation had a depth of 20 m and had to remain open for
a period of up to 6 months. This was an uncommon problem that was complicated by
the presence of horizontal tectonic shear surfaces at depths of 13 to 15 m, and could
hardly have been examined without a sophisticated numerical model and some
calibration. The authors used the numerical model ICFEP used by Potts et al. (1997)
and similar soil parameters. The model and the input parameters were calibrated
against the failure of 2 cuts that occurred a few kilometers away before being applied
Page 14
to the Heathrow Airports Terminal 5 case. Finally, it was decided to accept a design
that was giving a calculated deep-seated failure in 1.42 years.
Rapid drawdown. Water bodies may apply pressure on totally or partially submerged
slopes and, if the level of water rapidly decreases (rapid drawdown), there are two
effects: a reduction of the stabilizing pressure on the slope itself and a change in
internal pore water pressures. As a consequence, several failures of slopes and of earth
dams have occurred. In most cases, the problem is neither fully drained nor fully
undrained, and rapid drawdown is accompanied and followed by progressive swelling
of the slope soil and change in pore water pressure that are controlled by the
coefficient of swelling/consolidation of the soil. Pinyol et al. (2008 and 2011)
examined this problem, using the coupled hydro-mechanical model CODEBRIGHT.
Pinyol et al. (2008), in particular, simulated the behavior of the Glen Shira Dam that
was subjected to a drawdown of about 9 m in 4 days (Fig. 11). Figures 11b and 11c
show the measured and calculated piezometric heads at the location of the
piezometers. The agreement is quite good. Figure 11b shows in particular that in
Piezometer 5, at an approximate elevation of 11 m, there is still a positive pore water
pressure of about 10 kPa when the reservoir has been lowered from its initial elevation
on day 9 to elevation 5.2 m at the end of drawdown, on day 13.
Fig. 11. Cross section of Shira dam subjected to rapid drawdown with the
location of the piezometers 1 to 5 (a); comparison between measured and
calculated piezometric heads (b and c) (modified after Pinyol et al., 2008).
Flow liquefaction. Several fully coupled hydro-mechanical models have also been
developed for predicting the response of earth structures to earthquakes. It is the case
of the model UBCSAND, run in FLAC, (Byrne et al., 2004) that is expressed in
effective stresses and associates strength reduction to pore pressure increase.
Page 15
Naesgaard et al. (2006) and Byrne et al. (2006), in particular emphasized the
importance of water flow and pore water redistribution within the soil mass, and
satisfactorily simulated both in space and time the failure of the Lower San Fernando
dam that occurred in 1971 some 20 to 30 s after earthquake shaking (Fig. 12).
Fig. 12. (a) Cross-section of the Lower San Fernando dam after failure (after
Seed et al., 1973) and (b) Simulated cross-section of the same dam after shaking
(from Naesgaard et al., 2006).
As shown above by the presented examples, a large number of geotechnical
problems are coupled, and the new generation of numerical models allows their study.
It is also important to note that the combination of monitoring and numerical models
provides important means to verify the hypotheses on the mechanisms involved and
the representativeness of the model, and to calibrate the input parameters.
ASSESSMENT OF SLOPE STABILITY IN PRACTICE
The assessment of slope stability is difficult and requires an experienced geotechnical
engineer. The approach suggested here is not new as such but reflects the way the
authors and some of their colleagues follow or at least try to follow. It is summarized
in the flow chart below (Fig. 13). The main stages are: preliminary office work; site
visit; qualitative stability assessment; investigation; quantitative stability assessment;
post-failure characterization; and management options. Fell et al. (2000) present a
series of questions to be addressed in slope stability investigations and the reader is
encouraged to refer to that work.
Preliminary office work
Slope stability assessment generally starts with a preliminary office work. At that
stage, we are looking for:
A general idea of the geology of the site as well as geological and
Page 16
geomorphological features.
The geometry of the slope (height, average inclination and profile) and
possible changes in geometry due to human activity and/or erosion.
The geomaterials involved.
Surface waters (ponding, drainage, etc.).
Preliminary regional and local hydrogeological models.
Types of landslides that have occurred or may occur at the site and in the
vicinity, and, if information is available, their characteristics (type and stage of
movement; geometry; retrogression; runout distance).
Hypotheses on possible causes of movements.
Indicators of evidence of movement, evidence of previous landslides, evidence
of erosion at the toe, presence of cracks, vegetation, presence of debris, etc.
Local climatic and seismic conditions.
Documents that can be considered:

- If available, zonation maps.
- Topographic maps and LiDAR surveys.
- Historical data: witness reports; archives; newspapers; etc.
- Aerial photographs of the site and of its vicinity at different times for
stereographic observations. Their comparison may show the evolution of the
slope (erosion at the toe, surcharging at the top; changes in vegetation; etc.).
- Geological and geomorphological maps.
- Seismicity maps.
- Climatic conditions (rainfalls in particular).
- Previous investigations in the area (existing borings and soundings; laboratory
and in situ tests; pore pressure observations in pits, tubes or piezometers;
hydrogeological modeling; previous slope stability assessments).
The preliminary Office Work should be completed by a sub-report that should contain:
The results of the preliminary analysis of the available information, in
particular, the presupposed involved geomaterials, type of movement,
predisposition factors, and possible triggering or aggravating factors.
The additional information and parameters that may be required, and the type
of investigation that should be performed to complete the information
(boreholes, piezometers, inclinometers, etc.)
The elements that should be specifically examined during the site visit. A site
visit checklist could be used for that purpose.
Site visit
The preliminary office study being completed, a site visit is mandatory. There are a
number of aspects to look for or to examine:
- Location of the site in its geomorphological and climatic environment.
- Geomorphological features of the site.
- Recent history of the site, including changes in surface drainage and geometry.
Discussions with owners and neighbors are often helpful.
Page 17
Fig. 13 Flow chart for slope stability assessment

-
Evidences of geological processes such as landslides or erosion at the site and

in the vicinity. The characteristics of the landslides in the vicinity of the slope
(type, retrogression and runout distance of the debris) must be documented.
The characteristics of the erosion, if any, (location, abruptness and height of
the eroded zone) must also be documented.
Geomaterials constituting the slope, which can be observed in eroded zones or
along cuttings or scarps of previous landslides.
Evidences of precarious stability and movements: open cracks, fissures and
shears, possibly with vertical displacements; evidences of displacements;
Page 18
bulging of the slope; heaving near the toe; evidences of deformation of manmade works (fissures and cracks in walls, road pavements, buildings; opening
of joints in walls; tilting of walls or structures; etc.).
- Type of vegetation. If there are trees, describe the type, the approximate age
(possibility to use dendrochronology), and the verticality.
- Resurgence or evidence of wet areas on the slope.
- Geomaterial at the toe of the slope, possibly in a river, which can indicate
draining conditions in the slope.
- Access to the site for in situ investigation.
- Locations for possible soundings, boreholes and instruments such as
piezometers, inclinometers, etc.
- Identification of potential problems that could occur during the study.
The site visit must be documented in a preliminary sub-report.
Qualitative stability assessment
Many slopes have remained stable for centuries or for thousands of years, and have
already experienced extremely severe groundwater conditions. The probability of deep
failure in the future is thus extremely low. This has led the government of the Province
of Quebec to adopt an approach for the assessment of slope stability in sensitive clays
in which a qualitative assessment of stability is part of the methodology (Lefebvre et
al., 2008). This may lead to consider as stable slopes that possibly have a factor of
safety smaller than the value generally required (e.g. F > 1.5), but it is thought that this
approach based on performance is as reliable as the usual approach based on the
calculation of a factor of safety (Lefebvre et al., 2008). It is thought that such
qualitative assessment could be used in other geological environments. The conditions
for considering a slope stable in the future are expressed hereunder in terms that are
more general than those expressed by Lefebvre et al. (2008):
The slope must be in a homogeneous geological and geomorphological
context, thus with similar slopes in the vicinity.
The slope should have had the same geometry for a long period of time
(centuries), which means that it should not have been submitted to erosion and
thus to a progressive decrease of its factor of safety. If the geometry has been
recently modified by human activities, the changes must have a positive effect
on stability; if the effect on stability is negative (e.g. excavation at the toe or
loading at the top), even if minor, the slope cannot be a priori considered
stable.
The vegetation must not have changed for a long period of time. The presence
of mature forest is a good indicator of stability; if the slope has been
deforested, it generally has a negative effect on stability and the slope cannot a
priori be considered stable.
The slope geometry (height and slope angle) must be less critical than the
neighboring slopes (and thus gives some stability reserve) or known from local
experience to be generally stable.
There should be absence of previous landslides in the neighboring and similar
slopes, indicating that the area is not prone to landslides.
Page 19
If such criteria are fulfilled (at least in the Quebec context), the slope is considered
stable without any calculation of factor of safety. If there is uncertainty on the
conclusion, a preliminary investigation may be required for verifying that the local soil
conditions are not different from those assumed (e.g. by drilling boreholes or
performing piezocone tests) or that there is no excessive pore pressures under the toe
of the slope. Also, monitoring may help in confirming the stability of the considered
slope or may be used in an observational approach.
This approach should be used with caution (or may not be used) for shallow
landslides, for several reasons: (1) due to climate changes, slopes may be influenced
differently from what has existed in the past. In other words, the principle put forward
by Hutchinson (1995) that past and present are keys for the future may not be
always true; (2) there are areas in the world, in particular in tropical regions, where
weathering is active and may rapidly change the soil properties with time.
The qualitative stability assessment should also be summarized in a preliminary subreport.
Investigation and instrumentation
The investigation must aim at defining the characteristics of the slope (in particular the
potential predisposition factors), the characteristics of the geomaterials involved and
the hydrogeological conditions. It may also help, with some instrumentation, at
specifying the characteristics of the movement (e.g. existence of a pre-existing failure
surface; rate of movement) and the variations of pore water conditions.
The in situ investigation may consist of:
- Boreholes, trenches and pits (for observation and recovering of soil samples)
and soundings (piezocones and others);
- Geophysics: seismic P and S waves and resistivity for sub-aerial
investigations; multibeam echo-sounding and seismic reflection for offshore
investigations (Locat and Lee, 2002, 2009).
- Hydrogeology characterization: water levels in pits; piezometers; tensiometers
and psychrometers. The installation of these instruments, their reliability and
their monitoring have to be thought: number and location; possibility to have
artesian pressures; except at shallow depths, simple tubes should not be used
for pore water pressure measurements as the observations of water head can be
misleading, and only isolated piezometers should be used; choice of the best
instruments considering partial saturation, time lag, frequency of
measurements and type (manual or remote) of readings; duration of the
monitoring period. The hydraulic conductivity of the different soil units may
also have to be estimated or measured.
- Weather station that may have to be installed on the site or may exist at small
distance.
- Installation of instruments for monitoring.
In soils not containing gravels or coarser elements and in relatively soft soils, the
piezocone is a very useful tool as it provides a detailed stratigraphy of the deposit as
well as pore water pressure data after dissipation tests. In slopes in clay, the piezocone
may also show evidences of destructuration, and thus indications of precarious
Page 20
stability (Demers et al., 1999; Leroueil, 2001).

The laboratory investigation aims at determining physical characteristics (unit
weight; degree of saturation; grain size distribution; water retention curve; etc),
mechanical properties (strength parameters in particular) and hydraulic properties
(saturated hydraulic conductivity; hydraulic conductivity versus degree of saturation
for unsaturated soils; etc.). Some remarks can be made: it is difficult to obtain
undisturbed characteristics of soils that are heterogeneous or fissured; the size of
specimens has to be established as a function of grain size and intensity of fissuring;
most mechanical parameters are not intrinsic but depend on many factors (degree of
disturbance of the specimens; size of the specimen; type of test; stress conditions; rate
of testing; etc.), and this has to be thought before selecting testing conditions and the
parameters for stability analyses. In some cases, only in situ tests (as direct shear tests
or permeability tests in boreholes) can be performed to measure representative soil
properties.
At this stage, instrumentation aims mostly at specifying the potential type of
movement and its characteristics (e.g. inclinometers for specifying the stage of
movement, pre-failure movement or reactivation, and, in this latter case, the location
of the failure surface), the rate of movement and hydrogeological conditions.
If landslides have occurred in the vicinity in similar geological conditions, it may be
interesting to perform back analyses in order to define strength parameters (without
forgetting that they are associated with hypotheses on stratigraphy and pore pressures)
and compare them with those deduced from laboratory tests. In the case of soils that
cannot be easily sampled, back analyses may become the only way for assessing the
mobilized shear strength parameters.
The investigation must be associated with an evaluation of its quality. In particular,
the coherence of the data has to be checked: coherence of the different test results;
coherence between soil stratigraphy and properties; coherence between stratigraphy
and hydrogeological conditions. In relation with this latter point, it is important to
mention that errors on pore pressure measurements are common. Also, when possible,
it may be interesting and useful to try to correlate measured parameters such as
weather conditions, pore pressures and rate of movement.
For the quantitative stability assessment discussed below, it is important to think that
the worse conditions, in particular in terms of pore pressures, have to be considered.
A preliminary sub-report on the investigation performed and the results should be
prepared at that stage. It should present the geological and geotechnical model of the
slope with the geometry and the stratigraphy of the slope, and the hydraulic and
mechanical characteristics of the different soil units.
Quantitative stability assessment
The quantitative stability assessment concerns first-time failures or reactivation, the
post-failure stage being considered only if there is possibility of failure. It must be
performed for the worse predictable conditions in terms of pore pressure conditions,
changes in geometry due to erosion or human activity, earthquake activity, possibility
of rapid drawdown, etc.
Traditionally, this assessment has been established by the determination of a factor
Page 21
of safety in 2-D limit equilibrium analyses. In addition to the geometry of the slope
and the stratigraphy, these analyses require the unit weight and the strength parameters
of the different soil units. They also require the hydrogeological regime or pore
pressure distribution in the slope for the worse estimated conditions. Except for very
simple cases where the pore pressure distribution can be easily estimated, it is
recommended to use a hydrogeological model to define this distribution, remembering
that in natural slopes pore pressures are not hydrostatic. Also, if there are variations of
pore pressures at the boundaries, the variations at depth are not the same as at the
boundaries (see Fig. 8). For detailed analysis of such cases, coupled hydro-mechanical
models can be the only way to get reliable results.
For the selection of the strength parameters, when possible and for the case of brittle
soils, the best way is from back-analyses of past landslides; otherwise, the engineer
has to use his experience and judgment. In fact, if for ductile geomaterials, the strength
parameters measured in the laboratory are generally representative of in situ behavior,
it is not the case for brittle geomaterials where, mostly due to progressive failure and
strain rate effects, the peak strength measured in the laboratory on good quality
samples generally overestimates the strength mobilized in situ.
The required factor of safety for stable conditions depends on the country and/or
on the author, but is most of the time equal to 1.5 or 1.4. This factor of safety covers to
a large extent uncertainty on geometry, stratigraphy, strength parameters, pore
pressure conditions, soil behavior and calculation method. However, if it is considered
that the stability of a slope has to be improved, an increase in the existing factor of
safety by 20 to 25% is generally considered sufficient. Fell et al. (2000) mention that
for large landslides, an increase in F of 0.05 to 0.1 has been considered adequate.
It is worth noting that several countries are now using limit state methods of analysis
based on partial factors applied on soil parameters and actions. This is in particular the
case in Europe with the Eurocode 7 (EN 1997) mandatory since March 2010.
2-D Limit equilibrium methods of analysis present several shortcomings:
- They assume that the soil above the failure surface behaves as a rigid mass
without any consideration for its stress-strain behavior and that the factor of
safety is the same all along the failure surface. This is generally not the case.
- Most failure surfaces are three dimensional and several authors have examined
the effect of the third dimension on the factor of safety (e.g. Skempton, 1985,
Leshchinsky and Huang, 1992, and Morgenstern, 1992). According to
Skempton (1985), it would approximately increase the factor of safety
calculated in 2-D analyses by (1 + KD/B), where D and B are the average
depth and width of the sliding mass and K is an earth pressure coefficient.
However, as 3-D analyses are more complex and less conservative than 2-D
analyses, their use is not spread in the profession.
- As mentioned by Tavenas et al. (1980), when the analyses are performed in
effective stresses, the definition of the factor of safety implicitly assumes a
specific stress path that may be quite different from the stress path really
followed. In a stress diagram such as the one shown in Figure 9b, the implicitly
assumed stress path is approximately as IG whereas the stress path followed
when failure is reached by pore pressure increase may be as IYC, thus quite
different.
Page 22
There is a priori no consideration for the consequences of failure; in particular,

the characteristics of the post-failure stage (velocity and runout distance) are
not considered.
In spite of these limitations, limit equilibrium analyses have been very useful to the
profession and are still widely used. It has to be remembered however that in case of
soils with a strain softening behavior, the average mobilized strength is lower than
the peak strength measured in the laboratory on good quality samples. In most
geological environments, analysis of the stability of slopes by limit equilibrium
method has thus to be seen as a semi-empirical approach, and, when possible, the back
analysis of past failures in the same geological context can be used to specify strength
parameters.
Factor of safety and probability of failure are often thought to be related: if F is close
to 1.0, the probability of failure is considered very high; if it is about 1.5, the
probability of failure is thought to be extremely small. However, as shown by several
researchers (Christian et al., 1994; Duncan, 2000; Nadim et al., 2005), there may be no
direct correspondence between factor of safety and probability. There have also been
attempts for using concepts of probability and risk (Eq. 1) in the last decades; in
particular, a specialty Conference on Landslide Risk Management was held at
Vancouver in 2005. Picarelli et al. (2005) notice the difficulty in quantifying hazard
and that a good understanding of the processes involved is an essential step to any
quantification of hazard. There are several reasons that can explain the difficulties in
quantifying hazard: the variability of soil units and soil characteristics; the uncertainty
on the worse conditions to be considered; the uncertainty on the terrain model; and the
uncertainty on the considered mechanical soil model. Nadim et al. (2005) indicate that
probabilistic analyses can be seen as a complement to the conventional deterministic
safety factor.
There are possibilities to consider the risk in deterministic approaches through the
selected required factor of safety. Smaller values could be accepted for events with
small consequences or small probability of occurrence. Along these lines: in 1984 in
Hong Kong (GEO, 1984), the recommended F value was increasing with the potential
loss (economic and life); in Sweden, Sllfors et al. (1996) indicate required factors of
safety that decrease when the use of land changes from New development (high
consequences) to Undeveloped land (low consequences), and when the quality of
the investigation increases (lower uncertainty).
Factors such as historic information, geomorphological information and climaticlandslide activity relationships should also be considered for assessing the stability of
slopes (Leroueil & Locat, 1998). Monitoring data, possibly in combination with
numerical analyses, may also play an important role in the assessment of slope
stability by specifying the slope conditions. Because of all these aspects that deserve
consideration, the authors do not think that a single and fixed value of requested factor
of safety should be considered for assessing the stability of slopes.
A way to overcome most of the limitations of limit equilibrium methods is to use
finite element/finite difference methods, or possibly discrete element methods for
some cases of rock slopes, which have been developed and used in the context of
slopes (Potts et al., 1997; Griffiths and Lane, 1999; Eberhardt et al., 2004; Byrne et al.,
2006). These methods present a number of advantages: they do not require
Page 23
assumptions on the shape and location of the failure surface; through the constitutive
equations (that may be valid in saturated and unsaturated conditions), they provide
information on deformation; they may allow the development of progressive failure; if
there is hydro-mechanical coupling, they also allow pore water and pore pressure
redistribution within the soil mass; they give detailed information on the processes
involved and thus a better understanding of slope behavior. They also give the
possibility to assess slopes in terms of their behavior (mostly acceptable deformations)
rather than in terms of an obscure factor of safety. If their use has mostly been limited
to research and the analysis of a few complex cases, it is certain that they will be more
and more often used in practice in the years to come.
Special case: Cuts and rapid drawdown. These are problems for which changes of
effective stresses, and thus in factor of safety, depend on the consolidation/swelling
process of soil mass and thus on time. Except when the process is fully drained or
fully undrained, the most relevant approaches for analyzing the evolution of stability
consist in using numerical coupled hydro-mechanical models. The works done by
Kovacevic et al. (2007) and Pinyol et al. (2008 and 2011 (for the analysis of the
Canelles landslide)) and mentioned above are good examples of applications. Other
approaches have been developed and are summarized by Cornforth (2005) and
Duncan and Wright (2005).
Special case: Seismic slope stability analyses. Earthquakes induce cyclic
accelerations, mostly horizontal, and thus shear stresses into the soil mass. They result
in dynamic forces but also possible increase in pore pressure, destructuration of the
soil, and consequently reduction in the shear strength of the soil. Consequences may
be landslides or large movements, possibly with soil liquefaction or development of
permanent deformations, or reactivation of landslides. It is then necessary to verify the
slope response to these conditions. It is however not easy as the processes involved are
very complex and the state of practice is still in evolution. The reader may refer to
Blake et al. (2002) and Power et al. (2006) for guidelines and additional references.
Essentially, it is important to differentiate between soils that cannot lose significant
strength due to earthquake loading (ductile soils) and those that can lose strength and
liquefy (brittle soils, in drained or undrained conditions).
In the first category, the most common approach is to perform a pseudo-static limit
equilibrium analysis in which a destabilizing force equal to the soil weight multiplied
by a seismic coefficient is applied horizontally; if the calculated factor of safety is
lower than one, soil displacements arising from shaking can be determined on the
basis of Newmark (1965)s approach that considers that the mass of soil moves when
the yield acceleration is exceeded (see Blake et al., 2002, and Power et al., 2006).
For soils that may show a strain softening behavior or liquefy, several steps are
usually followed: the first one is to check if the earthquake may trigger soil
liquefaction (Youd et al., 2001; Idriss and Boulanger, 2008); if so, the second one
consists in verifying if the post-liquefaction strength (Seed and Harder, 1990; Olson
and Stark, 2002) may prevent a flow-slide. Displacements can be determined on the
base of Newmark (1965)s approach (see Youd et al., 2002; Idriss and Boulanger,
2008).
Page 24
Another but more complex approach would be to use a dynamic fully coupled
hydro-mechanical numerical model incorporating constitutive equations expressed in
effective stresses that associate strength reduction to pore pressure increase (e.g.
DYNAFLOW (Prevost, 2002); UBCSAND (Byrne et al., 2004; Naesgaard et al.,
2006).
This important stage of quantitative slope stability assessment should be completed
with the preparation of a preliminary sub-report.
Post-failure characterization
The post-failure stage has to be examined separately from the failure stage as it
corresponds to a completely different process. At the time of failure, some potential
energy (EP) becomes available, and what happens then depends on how this energy is
redistributed. Part of the potential energy will dissipate through friction (EF); the rest
will be dissipated in breaking up, disaggregating and remolding the soil (ED) and for
generating movement (kinetic energy, EK). Over a time interval during the post-failure
stage:
'EP + 'EF + 'ED + EK = 0
[2]
Leroueil (2001) discusses the different components of Equation 2 and their practical
implications. The disaggregating energy in chalk and rocks in general has been
examined by Leroueil (2001) and Locat et al. (2006) respectively; the remolding
energy of sensitive clays has been examined by Leroueil et al. (1996) and Locat et al.
(2008). Equation 2 also implies that in ductile materials, the available potential energy
is dissipated through friction and, consequently, kinetic energy and post-failure rates
of movement are small. On the other hand, if the material shows a loss of shear
strength after failure, part of the available potential energy goes into movement (E K)
and high velocities as well as long runout distances could be reached.
As at the failure stage, it is essential to understand the mechanical process to predict
the post-failure stage. For example, to have a flow-slide in sensitive clays, it is
necessary to generate enough potential energy during the first-time failure to remold
the clay, and to have clay with a liquidity index larger than 1.0 so that it can flow once
remolded (Tavenas, 1984). Olivares and Damiano (2007) examined the post-failure
mechanisms of initially unsaturated shallow deposits of pyroclastic soils on steep
slopes from the Campania Region, Italy. These deposits are susceptible to landslides
due to rainwater infiltration. From field and laboratory observations, and the
understanding of unsaturated soil mechanics, Olivares and Damiano (2007) came to
the following conclusions: (1) in steep slopes with a slope angle near or slightly larger
than the friction angle I of the soil, this latter is practically saturated at the onset of
failure and flow-slide can develop; (2) in very steep slopes with a slope angle
significantly larger than I, the soil at the time of slope failure is far from complete
saturation and the possibility of having a flow-slide is less likely. These kinds of
information can be used in a qualitative assessment of the possibility of flow-slide.
Post-failure assessment is important as it is often the most destructive stage of the
landslide since the materials involved reach their maximum velocity, and their largest
Page 25
displacement. In many cases where the debris stops close to the toe of the slope, the
runout distance can be defined on the basis of local observations and experience.
However, if a flow-like landslide is expected, it is much more complex as the process
depends on the geomaterial involved and its physical characteristics, and on the
topography and other characteristics (materials that can be entrained, vegetation, etc.)
of the travel path. Numerous numerical models have been developed for flow-like
landslides (DAN3D, SHWCIN, RASH3D, TITAN2D, DFEM, FLO-2D, etc.) and a
review is made by McDougall et al. (2008). The main difficulty with these models is
the selection of input parameters that can provide representative bulk behavior (travel
distance, velocities, thickness of debris). Most of the time, these parameters cannot be
measured in the laboratory and result from trial-and-error adjustments of the
simulation of previous events. McDougall et al. (2008) also discuss these aspects. To
get more information about these special problems, the reader can refer to the
Proceedings of the International Workshop on Occurrence and Mechanisms of Flowlike Landslides in Natural Slopes and Earthfills, Sorrento, Italy, 2003 and of the
International Forum on Landslide Disaster Management, Hong Kong, 2007; the first
meeting considered experiences from a variety of places in the world, while the
second one included a special session where some benchmark cases were analyzed by
different researchers, using a number of numerical codes.
Post-failure may include the triggering of a tsunami, as in some cases of rock falls or
rockslides along Norwegian fjords (Lacasse et al., 2008) or of coastal or submarine
landslides.
If the study of post-failure is relevant, this stage should also be the object of a
preliminary sub-report.
Defining management options
For defining management options, it is essential to well understand the mechanical
processes that could lead to failure or large movements, and to know the
characteristics of the potential landslide (size, velocity, travel distance).
Risk assessment and Management is the topic of an invited lecture to this
GeoCongress, by Suzanne Lacasse, and only a few elements are provided here
concerning management options. However, it is thought that, even without fully
considering the risk, the engineer who concludes of the precarious stability of a slope
must suggest possible options for mitigating the risk (Eq. 1). The general options are:
- Avoid the problem and choose a different lay-out or site.
- Decrease the hazard by reducing the driving forces or increasing the resisting
forces or doing both (see Table 1). Mitigation methods and their design are not
examined in this Paper. The reader is referred to Cornforth (2005) and/or
LCPC (1998) for the selection and design of remedial and prevention methods.
The mitigation methods are often associated with some monitoring in order to
verify the hypotheses made and the efficiency of the adopted system.
- Decrease the risk by building in risk areas passive works, as check dams,
barriers or retention basins, in order to protect the elements at risk (Versace et
al. 2009).
- In some cases, mitigation measures cannot be applied due to economic,
Page 26
morphological or technological constraints. The consequences of a potential

event and thus the associated risk can then be reduced by monitoring rainfalls,
or pore water pressures, or rates of displacement, or acoustic emission, etc.
(what Picarelli et al. (2011) describe as geo-indicators) to define time close to
failure at which planned measures can be taken for the protection of people and
goods or at which people can be evacuated. This time can be defined on the
basis of experience (e.g. antecedent rainfall and rainfall intensity exceeding
some values), calculated factor of safety (e.g. pore pressures exceeding given
values), or imminence of failure as indicated by the acceleration of
displacements, or excessive acoustic emission. This can be seen as an
observational approach or as an early warning system (Fell et al., 2000; Nadim
and Intrieri, 2011; Picarelli et al., 2011). In such approaches, numerical models
can be used for helping predicting and interpreting the measurements of the
geo-indicators.
As for the other stages of the analysis, the management options should also be the
object of a preliminary sub-report.
Final report
Preparing the final report on slope stability assessment is extremely important as it
implies the responsibility of the geotechnical engineer involved. It thus has to be done
in a very rigorous manner. The outline could coarsely follow the different stages
indicated in Figure 13 and described above, and the preliminary sub-reports suggested
for these different stages could be used to form a first draft. In order not to forget
apparent details that could then be of extreme importance, it is important to write these
preliminary reports as the project progresses.
CONCLUSION
Assessing the stability of slopes is a very difficult task that requires an experienced
engineer and a rigorous approach. For that purpose, the geotechnical characterization
of slopes that classifies the information on slopes is a very useful tool.
The paper evidences the tremendous progresses that have been made in the domain
of slopes: there are new technologies, with possibilities to take automatically readings
and send them via wireless links to an interpretation center; there are also interesting
progresses in our understanding of some aspects of soil and slope behaviors; there has
been development of very powerful coupled hydro-mechanical numerical models that
can consider saturated and unsaturated soils, progressive failure, etc., and allow the
analysis of complex problems. All this constitutes a real revolution in our way to
monitor and analyze slopes.
A methodology, not new as such, is proposed for the assessment of slope stability.
Its main characteristics are as follows:
- It includes a qualitative stability assessment based on past performance and on
comparison with slopes in the vicinity that have similar geomorphology.
- It recognizes the shortcomings of limit equilibrium analyses but considers that
they have been extremely useful to the profession and will remain in use in
Page 27
practice.
It is thought that the required factor of safety should not have a fixed value, but
a value that integrates the concepts of risk, increasing with the consequences of
a potential failure, decreasing when the uncertainty decreases and when the
understanding of the processes involved progresses; also, the value could be
reduced if associated with an adequate monitoring and an observational
approach. The authors however recognize the difficulties for establishing
guidelines for this approach.
- Because the processes involved are completely different from those at the
stages of failure and reactivation, and because its consequences may be major,
it is considered that the post-failure stage should be examined systematically
and separately.
The coupled hydro-mechanical numerical models overcome most of the shortcomings
of the limit equilibrium methods and give the possibility to assess slopes in terms of
their behavior (mostly deformations) rather than in terms of factor of safety. There is
no doubt that they will be more and more often used in practice in the future.
ACKNOWLEDGEMENTS
The authors acknowledge the contribution of colleagues from the Quebec Ministry of
Transports, Denis Demers in particular, and of Denis LeBoeuf, Guy Lefebvre, Ernest
Naesgaard, Catherine Cloutier, Luca Comegna and Ariane Locat. They also thank the
two anonymous reviewers for their constructive comments.
REFERENCES
Amitrano, D., Arattano, M., Chiarle, M., Mortara, G., Occhiena, C., Pirulli, M. and
Scavia, C. (2010). Microseismic activity analysis for the study of the rupture
mechanisms in unstable rock masses. Natural Hazards and Earth System Sciences,
10: 831-841.
Ashford, S.A. and Sitar, N. (1994). Seismic Response of Steep Natural Slopes. Report
No. UCB/EERC-94/05, Earthquake Engineering Research Center, University of
California at Berkeley.
Baum, R.L., Godt, J.W., Harp, E.L., McKenna, J.P. and McMullen, S.R. (2005). Early
warning of landslides for rail traffic between Seattle and Everett, Washington,
U.S.A. Proc. Int. Conf. on Landslide Risk Management, Vancouver: 731-740.
Bernander, S. (2000). Progressive landslides in long natural slopes, formation,
potential extension and configuration of finished slides in strain-softening soils.
Licentiate Thesis, Lulea University of Technology, Lulea, Sweden.
Berz, D., Boore, G., Bouwkamp, D., Hackenbeck, J., McGuire, U., Sims, R., and
Wieczorek, G. (1980). Reconnaissance Report, Monte Negro, Yougoslavia
Earthquake, April, 15, 1979. Earthquake Engineering Research Institute, Berkeley,
California.
Blake, T.F., Hollingsworth, R.A. and Stewart, J.P. (2002). Recommended procedure
for implementation of DMG special publication 117 guidelines for analyzing and
mitigating landslide hazards in California. Document published by the Southern
Page 28
California Earthquake Center.

Brooks R. H. and Corey A.T. (1964). Hydraulic properties of porous media.
Hydrology Paper N.3, Colorado State Univ., Fort Collins, Colorado.
Byrne, P.M., Park, S., Beaty, M., Sharp, M., Gonzales, L. and Abdoun, T. (2004).
Numerical modeling of liquefaction and comparison with centrifuge tests. Canadian
Geotech. J., 41: 193-211.
Byrne, P.M., Naesgaard, E. and Seid-Karbasi, M. (2006). Hardy Lecture - Analysis
and design of earth structures to resist seismic soil liquefaction. Proc. 59th
Canadian Geotec. Engng. Conf., Vancouver, p. 1-24.
Cai, F. and Ugai, K. (2004). Numerical analysis of rainfall effects on slope stability.
Int. J. of Geomechanics, 4(2): 69-78.
Chandler, R.J. (1984). Recent European experience of landslides in over-consolidated
clays and soft rocks. Proc. 4th Int. Symp. on Landslides, Toronto, 1: 61-81.
Christian, J.T., Ladd, C.C. and Baecher, G.B. 1994. Reliability applied to slope
stability analysis. ASCE J. of Geotechnical Engineering, 120(12): 2180-2206.
Chu, J., Leroueil, S. and Leong, W.K. (2003). Unstable behaviour of sand and its
implication for slope stability. Canadian Geotech. J., 40: 873-885.
Cloutier, C., Locat, J., Lord, P.-E. and Couture, R. (2011). Analysis of one year of
monitoring data for the active Gascons rockslide, Gasp Peninsula, Quebec. 5th
Canadian Conf. on Geohazards, Kelowna, B.C., Canada.
Collins, B.D. and Znidarcic, D. (2004). Stability analyses of rainfall induced
landslides. J. Geotech. and Geoenv. Engng, ASCE, 130(4): 362-372.
Comegna L., Guida A., Damiano E., Olivares L., Greco R. and Picarelli L. (2011).
Monitoraggio di un pendio naturale in depositi piroclastici sciolti. Proc. XIV
Convegno Nazionale di Geotecnica, Napoli, 2: 681-686.
Cornforth, D.H. (2005). Landslides in Practice. John Wiley & Sons, 596 p.
Couture, R., Charbonneau, F., Singhroy, V., Murnaghan, K., Drouin, H., Locat, J.,
Lord, P.-E. and Cloutier, C. (2011). PTA-InSAR rock slope monitoring at the
Gascons site, Gasp Peninsula, Quebec : Preliminary results. 5th Canadian Conf.
on Geohazards, Kelowna, B.C., Canada.
Cruden, D.M. and Varnes, D.J. (1996). Landslide types and processes. In Landslidesinvestigation and mitigation, Special Report 247, pp. 36-75. Washington,
Transportation Research Board.
Cruden, D.M., Krauter, E., Beltram, L., Lefebvre, G., Ter-Stepanian, G.I. and Zhang,
Z.Y. (1994). Describing landslides in several languages: the multilingual landslide
glossary. Proc. 7th Int. Congress of the Int. Association of Engng. Geology, Lisbon,
3: 1325-1333.
Damiano, E., Mercogliano, P., Olivares, L., Picarelli, L., Savastano, V., Schiano, P.
and Sikorski, B. (2012). A simulation chain from meteorological forecast to slope
instability in pyroclastic deposits. In preparation.
Damiano, E. (2003). Meccanismi di innesco di colate di fango in terreni piroclastici.
Ph.D Thesis, Seconda Universit di Napoli, Italy.
Damiano, E. and Olivares, L. (2010). The role of infiltration processes in steep slopes
stability of pyroclastic granular soils: laboratory and numerical investigation.
Natural Hazards, 52(2): 329-350.
Damiano, E., Olivares, L. and Picarelli, L. (2011). Steep slope monitoring in
Page 29
unsaturated pyroclastic granular soils: laboratory and numerical investigations.

Engineering Geology, submitted for publication.
DElia B., Esu F., Pellegrino A. and Pescatore T.S. (1985). Some effects on natural
slope stability induced by the 1980 Italian earthquake. Proc. 11th Int. Conf. on Soil
Mech. And Found. Engng., S. Francisco, 4:1943-1949.
Demers, D., Leroueil, S. and D'Astous, J. (1999). In situ testing in a landslide area at
Maskinong, Qubec. Canadian Geotechnical J., 36(6): 1001-1014.
Dixon, N., Hill, R. and Kavanagh, J. (2003). Acoustic emission monitoring of slope
instability: development of an active waveguide system. Geotechnical Engineering,
156(2): 83-95.
Dixon, N. and Spriggs, M. (2007). Quantification of slope displacement rates using
acoustic emission monitoring. Canadian Geotech. J., 44: 966-976.
Duncan, J.M. 2000. Factors of safety and reliability in geotechnical engineering.
ASCE J. of Geotechnical and Geoenvironmental Engineering, 126(4): 307-316.
Duncan, J.M. and Wright, S.G. (2005). Soil strength and slope stability. John Wiley &
Sons, 297 p.
Eberhardt, E., Stead, D. and Coggan, J.S. (2004). Numerical analysis of initiation and
progressive failure in natural rock slopes the 1991 Randa rockslide. Int. J. Rock
Mechanics and Mining Sciences, 41(1): 69-87.
Eberhardt, E., Watson, A.D. and Loew, S. (2008). Improving the interpretation of
slope monitoring and early warning data through better understanding of complex
deep-seated landslide failure mechanisms. Proc. 10th Int. Symp. on Landslides and
Engineered Slopes., Xian,1: 39-51.
Eurocode 7 EN 1997 Geotechnical Design (www.Eurocode-resources.com).
Fell, R., Hungr, O., Leroueil, S. and Riemer, W. (2000). Keynote Lecture
Geotechnical engineering of the stability of natural slopes, and cuts and fills in soil.
GeoEng2000 An Int. Conf. on Geotechnical and Geological Engineering,
Melbourne, 1: 21-120.
Fenelli, G.B., Picarelli, L. and Silvestri, F. (1992). Deformation process of a hill
shaken by the Irpinia earthquake in 1980. Proc. French-Italian Conf. on Slope
Stability in Seismic Areas, Bordighera, 47-62.
Fredlund D.G. and Rahardjo H. (1993). Soil Mechanics for Unsaturated Soils. WileyInterscience Publication, John Wiley & Sons, Inc.
Froese, C.R. and Moreno, F. (2007). Turtle Mountain Field Laboratory (TMFL): Part
1-Overview and activities. Proc. 1st North American Landslide Conference, Vail.
CD-Rom.
Gens, A. and Alonso, E.E. (2006). Aznalcollar dam failure. Part 2: Stability conditions
and failure mechanism. Gotechnique, 56(3): 185-201.
GEO (Geotechnical Engineering Office) (1984). Geotechnical Manual for Slopes, 300
p.
GEO-SLOPE Int. Ltd. (2004). Seepage model-ling with SEEP/W, users guide version
6.16. GEOSLOPE Int. Ltd., Calgary.
Greco, R. (2006). Soil water content inverse profiling from single TDR waveforms. J.
Hydrology, 317: 325-339.
Greco, R., Guida, A., Damiano, E. and Olivares, L. (2010). Soil water content and
suction monitoring in model slopes for shallow flowslides early warning
Page 30
applications. Physics and Chemistry of the Earth, 35: 127-136.

Griffiths, D.V. and Lane, P.A. (1999). Slope stability analysis by finite elements.
Gotechnique, 49(3): 387-403.
Hutchinson, J.N. (1995). Keynote paper: Landslide hazard assessment. Proc. 6th Int.
Symp. Landslides, Christchurch, 3: 1805-1841.
Idriss, I.M. and Boulanger, R.W. (2008). Soil Liquefaction during Earthquakes.
Report MNO-12, Earthquake Engineering Research Institute.
Ishihara, K. (1993). Liquefaction and flow failure during earthquakes. Gotechnique,
43(3): 351-415.
Ishihara, K. (2005). Characteristics of waterfront landslides induced by earthquakes.
Proc. Int. Conf. on Fast Slope Movements. Prediction and prevention for risk
mitigation, Naples, 2: 7-30.
Keefer D.K. (1984). Landslides caused by earthquakes. Bull. Geol. Soc. of America,
95: 406-421.
Kenney, T.C. and Lau, K.C. (1984). Temporal changes of groundwater pressure in a
natural slope of non-fissured clay. Canadian Geot. J., 21(1): 138-146.
Kokusho, T. (1999). Water film in liquefied sand and its effect on lateral spread.
ASCE J. Geotechnical and Geoenvironmental Engng, 125(10): 817-826.
Kokusho, T. (2003). Current state of research on flow failure considering void
redistribution in liquefied deposits. J. Soil Dynamics and Earthquake Engng., 23:
585-603.
Kovacevic, N., Hight, D.W. and Potts, D.M. (2007). Predicting the stand-up time of
temporary London clay slopes at Terminal 5, Heathrow Airport. Gotechnique,
57(1): 63-74.
Kulasingam, R., Malvick, E.J., Boulanger, R.W. and Kutter, B.L. (2004). Strength loss
and localization at silt interlayers in slopes of liquefied sand. ASCE J. Geotechnical
and Geoenvironmental Engng., 130(11): 1192-1202.
Lacasse, S., Eidsvig, U., Nadim, F., Hoeg, K. and Blikra, L. H. (2008). Event tree
analysis of Aknes rock slide hazard. Proc. 4th Canadian Conf. on Geohazards,
Quebec, 551-558.
Lade, P.V. (1993). Initiation of static instability in the submarine Nerlerk Berm.
Canadian Geotech. J., 30(6): 895-904.
LCPC (1998). Guide technique Stabilisation des glissements de terrain. Laboratoire
Central des Ponts et Chausses, Paris. 97 p.
Lefebvre, G., Demers, D., Leroueil, S., Robitaille, D. and Thibault, C. (2008). Slope
stability evaluation: more observation and less calculation. Proc. 4th Canadian Conf.
on Geohazards, Quebec, 413-420.
Lemke, R.W. (1966). Effects of the earthquake of March, 27, 1964, at Seward, Alaska.
Geological Survey Professional Paper 542-E
Leroueil, S. (2001). Natural slopes and cuts: movement and failure mechanisms.
Gotechnique, 51(3): 197-243.
Leroueil, S. (2004). Geotechnics of slopes before failure. Proc. 9th Int. Symp. on
Landslides, Rio de Janeiro, 2: 863-884.
Leroueil, S., Vaunat, J., Picarelli, L., Locat, J., Faure, R. and Lee, H. (1996). A
geotechnical characterization of slope movements. Proc. 7th Int. Symp. on
Landslides, Trondheim, 1: 53-74.
Page 31
Leroueil, S. and Locat, J. (1998). Slope movements: geotechnical characterization,

risk assessment and mitigation. Proc. 8th Congress Int. Assoc. Engng. Geology,
Vancouver, 933-944.
Leroueil, S., Chu, J. and Wanatowski, D. (2009). Slope instability due to pore water
pressure increase. Proc. First Italian Workshop on Landslides - Rainfall-induced
landslides, Naples, 1: 81-90.
Leshchinsky, D. and Huang, C.C. (1992). Generalized three-dimensional slopesstability analysis. ASCE J. of Geotechnical Engineering, 118(11): 1748-1764.
Lin, M.L., Wang, K.L. and Kao, T.C. (2008). The effects of earthquake on landslides A case study of Chi-Chi earthquake, 1999. Proc.10th Int. Symp. on Landslides,
Xian, 1: 193-201.
Locat, A. (2012). Ph.D. in preparation, Universit Laval, Qubec.
Locat, A., Leroueil, S., Bernander, S., Demers, D., Jostad, H.P. and Ouehb, L. (2011).
Progressive failures in eastern Canadian and Scandinavian sensitive clays.
Canadian Geotech. J. 48: 1696-1712.
Locat, J. (2011). La localisation et la magnitude du sisme du 5 fvrier 1663
(Charlevoix) revues laide des mouvements de terrain. Canadian Geotech. J.,
48(8): 1266-1286.
Locat, J., and Lee, H.J. (2002). Submarine landslides: advances and challenges.
Canadian Geotech. J., 39: 193-212.
Locat, J., and Lee, H.J. (2009). Submarine mass movements and their consequences:
An overview. K. Sassa, P. Canuti (eds.), Landslides Disaster Risk Reduction,
Springer-Verlag, (ch. 6) pp.: 115-142.
Locat, P., Couture, R., Leroueil, S., Locat, J. and Jaboyedoff, M. (2006).
Fragmentation energy in rock avalanches. Canadian Geotech. J. 43(8): 830-851.
Locat, P., Leroueil, S. and Locat, J. (2008). Remaniement et mobilit des debris de
glissements de terrain dans les argiles sensibles de lest du Qubec. Proc. 4 th
Canadian Conf. on Geohazards, Quebec, pp. 97-106.
Lu, N. and Godt, G. (2008). Infinite slope stability under steady unsaturated seepage
conditions. Water Resources Research, 44 W11404.
Malvick, E.J., Kutter, B.L. and Boulanger, R. (2008). Postshaking shear strain
localization in a centrifuge model of a saturated sand slope. ASCE J. of
Geotechnical and Geoenvironmental Engineering, 134(2): 164-174.
McDougall, S. and Hungr, O. (2004). A model for the analysis of rapid landslide
motion across three-dimensional terrain. Canadian Geotech. J. 41: 1084-1097.
McDougall, S., Pirulli, M., Hungr, O. and Scavia, C. (2008). Advances in landslide
continuum dynamic modelling. Proc. 10th Int. Symp. on Landslides and Engineered
Slopes, Xian, 1: 145-157.
Morgenstern, N.R. 1992. Keynote Paper: The role of analysis in the evaluation of
slope stability. Proc. 6th Int. Symp. on Landslides, Christchurch, 3: 1615-1629.
Nadim, F., Einstein, H. and Roberds, W. 2005. Probabilistic stability analysis for
individual slopes in soil and rock. Proc. Int. Conf. on Landslide Risk Management,
Vancouver, pp. 63-98.
Nadim, F. and Intrieri, E. (2011). Early warning systems for landslides. Challenges
and new monitoring technologies. 5th Canadian Conf. on Geohazards, Kelowna,
Canada, on CD.
Page 32
Naesgaard, E., Byrne, P.M. and Seid-Karbasi, S.M. (2006). Modeling flow
liquefaction and pore water pressure redistribution mechanisms. Proc. 8th U.S.
National Conf. Earthquake Engng., San Francisco, on CD.
Newmark, N.M. (1965). Effects of earthquakes on dams and embankments.
Gotechnique, 15(2): 139-160.
Ng, C.W.W. and Shi, Q. (1998). A numerical investigation of the stability of
unsaturated soil slopes subjected to transient seepage. Computer and Geotechnics,
22(1): 1-28.
Ochiai, H., Okada, Y., Furuya, G., Okura, Y., Matsui, T., Sammori, T., Terajima, T.,
and Sassa, K. (2004). A fluidized landslide on a natural slope by artificial rainfall.
Landslides, 1(3), 211-220.
Olivares L. (1997). Caratterizzazione dellargilla di Bisaccia in condizioni monotone,
cicliche e dinamiche e riflessi sul comportamento del Colle a seguito del
terremoto del 1980. PhD Thesis, Universit di Napoli Federico II
Olivares, L. and Damiano, E. (2007). Postfailure mechanisms of landslides: laboratory
investigation of flowslides in pyroclastic soils. ASCE J. of Geotechnical and
Geoenvironmental Engineering, 133(1): 51-62.
Olivares, L. and Tommasi, P. (2008). The role of suction and its changes on stability
of steep slopes in unsaturated granular soils. Proc. 10th Int. Symp. on Landslides,
Xian, 1: 203-215.
Olivares, L., Damiano, E., Greco, R., Zeni, L., Picarelli, L., Minardo, A., Guida, A.
and Bernini, R. (2009). An instrumented flume to investigate the mechanics of
rainfall-induced landslides in unsaturated granular soils. Geotechnical Testing
Journal, 32(2): 108-118.
Olson, S.M. and Stark, T.D. (2002). Liquefied strength ratio from liquefaction flow
failure case histories. Canadian Geotech. J., 39: 629-647.
Papa, R., Pirone, M., Nicotera, M. and Urciuoli, G. (2009). Meccanismi di innesco di
colate di fango in piroclastiti parzialmente sature. Proc. First Italian Workshop on
Landslides - Rainfall-induced landslides, Naples, 2: 91-106.
Phillips, R. and Coulter, S. (2005). COSTA-C Centrifuge Test Data Report, C-CORE
Report R-04-082-075. Memorial University of Newfoundland, Canada.
Picarelli L., Versace P., Olivares L. and Damiano E. (2009). Prediction of rainfallinduced landslides in unsaturated granular soils for setting up of early warning
systems. Proc. Int. Forum on Landslide Disaster Management, Hong Kong, 1: 643665
Picarelli, L., Damiano, E., Greco, R., Olivares, L. and Zeni, L. (2011). Geoindicators
for timely prediction of fast precipitation-induced landslides in unsaturated granular
soils. Canadian Geotech. J., submitted for publication.
Picarelli L. and Zeni L. (2009). Discussion to Test on application of distributed fiber
optic sensing technique into soil slope monitoring by B.J. Wang, K. Lee, B. Shi
and J.Q. Wei. Landslides, 6(4): 361-363.
Picarelli, L., Oboni, F., Evans, S.G., Mostyn, G. and Fell, R. (2005). Hazard
characterization and quantification. Proc. Int. Conf. on Landslide Risk
Management, Vancouver, pp. 27-61.
Pinyol, N., M., Alonso, E.E. and Olivella, S. (2008). Rapid drawdown in slopes and
embankments. Water Resources Research 44, W00D03, 22 pp. Special issues on:
Page 33
Hydrology and Mechanical Coupling in Earth Sciences and Engineering:

Interdisciplinary Perspective.
Pinyol, N.M., Alonso, E.E., Corominas, J. and Moya, J. (2011). Canelles landslide.
Modelling rapid drawdown and fast potential sliding. Landslides, In press.
Potts, D.M., Dounias, G.T. and Vaughan, P.R. (1990). Finite element analysis of
progressive failure of Carsington embankment. Gotechnique, 47(5): 953-982.
Potts, D.M., Kovacevic, N. and Vaughan, P.R. (1997). Delayed collapse of cut slopes
in stiff clay. Gotechnique, 47(5): 953-982.
Power, M., Fishman, K., Makdisi, F., Musser, S., Richards, R. and Youd, L. (2006).
Seismic retrofitting manual for highway structures: Part 2-Retaining structures,
slopes, tunnels, culverts and roadways. MCEER-06-SP11, U.S. Department of
Transportation, Federal Highway Administration.
Prevost, J.H. (2002). Dynaflow A nonlinear transient finite element analysis
program, Version 2002. Release 01.A, Dept. of Civil Engng. and Operation
Research, Princeton University, Princeton, NJ.
Rahardjo, H., Lee, T.T., Leong, E.C. and Rezaur, R.B. (2005). Response of a residual
soil slope to rainfall. Canadian Geotech, J., 42: 340-351.
Rahardjo, H., Rezaur,, R.B., Leong, E.C., Alonso, E.E., Lloret, A. and Gens, A.
(2008). Monitoring and modeling of slope response to climate changes. Proc. 10th
Int. Symp. on Landslides, Xian, 1: 67-84.
Rollins, K., Herbst, M., Gerber, T. and Cummins, C. 2009. Monitoring displacement
vs depth in lateral pile load tests with shape accelerometer arrays. Proc. 17th Int.
Conf. on Soil Mechanics and Geotechnical Engineering, Alexandria, 3: 2016-2019.
Sllfors, G., Larsson, R. and Ottosson, E. (1996). New Swedish national rules for
slope stability analysis. Proc. 7th Int. Symp. on Landslides, Trondheim, 1: 377-380.
Saaltink, M.W., Ayora, C. and Olivella, S. (2005). User's guide for RetrasoCodeBright
(RCB).
Seed, R.B. and Harder, L.F. (1990). SPT-based analysis of cyclic pore pressure
generation and undrained residual strength. Proc. of the H.B. Seed Memorial
Symp., Bi-Tech Publishing, 2: 351-376.
Seed, H.B., Lee, K.L., Idriss, I.M. and Makdisi, F. (1973). Analysis of the slides of the
San Fernando Dams during the earthquake of 9 February 1971. Earthquake Engng.
Research Center, Report EERC 73-2.
Seid-Karbasi, M. and Byrne, P.M. (2007). Seismic liquefaction, lateral spreading, and
flow slides: a numerical investigation into void redistribution. Canadian Geotech.
J., 44: 873-890.
Shi, Y.-X., Zhang, Q. and Meng, X.-W. (2008). Optical fiber sensing technology used
in landslide monitoring. 10th Int. Symp. on Landslides, Xian, 1: 921-925.
Singhroy, V. and Molch, K. 2004. Characterizing and monitoring rockslides from
SAR techniques. Advances in Space Research, 33(3): 290-295.
Skempton, A.W. (1985). Residual strength of clays in landslides, folded strata and the
laboratory. Gotechnique, 35(1): 3-18.
Springman, S., Jommi, C, and Teysseire, P. (2003). Instabilities on moraine slopes
induced by loss of suction: a case history. Gotechnique, 53(1): 3-10.
Takahashi, C., Cai, F. and Ugai, K. (2008). FE analysis of performance of the Lower
and Upper San Fernando Dams under the 1971 San Fernando earthquake. Proc. 10th
Page 34
Int. Symp. on Landslides and Engineered Slopes., 2: 1455-1461.

Tavenas, F. (1984). Landslides in Canadian sensitive clays: a state-of-the-art. Proc. 4th
Int. Symp. on Landslides, Toronto, 1: 141-153.
Tavenas, F., Trak, B. and Leroueil, S. (1980). Remarks on the validity of stability
analyses. Canadian Geotech, J., 17(1): 61-73.
Towhata, I., Ishihara, K., Kiku, H., Shimizu, Y., Horic, Y., Irisawa, T. (2001).
Submarine slides and land settlements in coastal areas during Kocaeli Earthquake.
Lessons learned from recent strong earthquakes. Proc. Satellite Conf. on
Earthquake Geotechnical Engineering, 71-76.
Uuciuoli, G., Picarelli, L., and Leroueil, S. (2007). Local soil failure before general
slope failure. Geotechnical and Geological Engineering J., 25(1): 103-122.
van Genuchten M. Th. (1980). A closed-form equation for predicting the hydraulic
conductivity of unsaturated soil. Soil Sci. Soc. Am. J., 44: 615-628.
Varnes, D.J. (1978). Slope movements: types and processes. In Landslides: Analysis
and Control. Transportation Research Board Report 176, pp. 11-33.
Varnes, D.J. and the IAGE Commission on Landslides and other Mass Movements on
Slopes. (1984). Landslide hazard zonation: a review of the principles and practice.
Paris: UNESCO.
Vaunat, J., Leroueil, S. and Faure, R. (1994). Slope movements: a geotechnical
perspective. Proc. 7th Congress Int. Assoc. Engng. Geol., Lisbon, 1637-1646.
Versace P., Capparelli G. and Picarelli L. (2009). Landslide investigations. The Sarno
case. Proc. 2007 Int. Forum on Landslide Disaster Management, Hong Kong, 1:
509-533.
Wang F., Cheng Q., Highland M., Miyajima M., Wang H. and Yan, C. (2009).
Preliminary investigation of some large landslides triggered by the 2008 Wenchuan
earthquake, Sichuan Province, China, Landslides, 6(1): 47-54.
Wang, G. and Sassa, K. (2001). Factors affecting rain-induced flowslides in laboratory
flume tests. Gotechnique, 51: 587-599.
Youd, T.L., Idriss, I.M., Andrus, R., Aranjo, I., Castro, G., Christian, J., Dobry, J.,
Finn, L., Harder Jr., L., Hynes, H.M., Ishihara, K., Koester, J., Liao, S.S.,
Marcusson III, W.F., Martin, G., Mitchell, J.K., Moriwaki, Y., Power, M.S.,
Robertson, P.K., Seed, R.B., and Stokoe II, K.H. (2001). Liquefaction resistance of
soils: Summary Report from the 1996 NCEER and 1998 NCEER/NSF Workshops
on Evaluation of Liquefaction Resistance of Soils. ASCE J. Geotechnical and
Geoenvironmental Engng., 127: 817-833.
Youd, T.L., Hansen, C.M. and Barlett, S.F. (2002). Revised multi-linear regression
equations for prediction of lateral spread displacement. ASCE J. Geotechnical and
Geoenvironmental Engng., 128: 1007-1017.
Page 35

Assessment of Slope Stability: Serge - Leroueil@gci - Ulaval.ca

Uploaded by

Copyright:

Available Formats

Assessment of Slope Stability: Serge - Leroueil@gci - Ulaval.ca

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Assessment of Slope Stability: Serge - Leroueil@gci - Ulaval.ca

Uploaded by

Copyright:

Available Formats

Downloaded from ascelibrary.org by dml u on 03/08/14. Copyright ASCE. For personal use only; all rights reserved.

Assessment of Slope Stability

Geotechnical Engineering State of the Art and Practice

GEOTECHNICAL ENGINEERING STATE OF THE ART AND PRACTICE 123

Fig. 1. Schematic slope movement characterization (from Leroueil et al., 1996)

Geotechnical Engineering State of the Art and Practice

124 GEOTECHNICAL ENGINEERING STATE OF THE ART AND PRACTICE

Fig. 2. Material types considered in the slope movement characterization (from

Fig. 3. Different stages of slope movements (from Leroueil et al., 1996)

Geotechnical Engineering State of the Art and Practice

GEOTECHNICAL ENGINEERING STATE OF THE ART AND PRACTICE 125

Table 1. Common triggering or aggravating factors (Leroueil, 2004)

Increase in shear stress

Another concept that helps in analyzing engineering problems and that is

in which H is the hazard or the phenomenon occurrence probability; Ri (for i = 1 to n)

Geotechnical Engineering State of the Art and Practice

126 GEOTECHNICAL ENGINEERING STATE OF THE ART AND PRACTICE

Pore water content and pressure

Some remarks can be made in relation with new systems:

Geotechnical Engineering State of the Art and Practice

GEOTECHNICAL ENGINEERING STATE OF THE ART AND PRACTICE 127

al., 2011): It is a rope-like array of sensors and microprocessors that fits in a

Geotechnical Engineering State of the Art and Practice

128 GEOTECHNICAL ENGINEERING STATE OF THE ART AND PRACTICE

Geotechnical Engineering State of the Art and Practice

GEOTECHNICAL ENGINEERING STATE OF THE ART AND PRACTICE 129

Geotechnical Engineering State of the Art and Practice

130 GEOTECHNICAL ENGINEERING STATE OF THE ART AND PRACTICE

Fig. 6. Measurement of rainfall (a and e), suction (Tensiometers; c and g) and

Geotechnical Engineering State of the Art and Practice

GEOTECHNICAL ENGINEERING STATE OF THE ART AND PRACTICE 131

reflected instantaneously to the entire soil deposit. Changes require a

Geotechnical Engineering State of the Art and Practice

132 GEOTECHNICAL ENGINEERING STATE OF THE ART AND PRACTICE

of failure released into kinetic energy.

Fig. 8. Location of piezometers and annual changes of piezometric level at Wabi

Geotechnical Engineering State of the Art and Practice

GEOTECHNICAL ENGINEERING STATE OF THE ART AND PRACTICE 133

Geotechnical Engineering State of the Art and Practice

134 GEOTECHNICAL ENGINEERING STATE OF THE ART AND PRACTICE

Geotechnical Engineering State of the Art and Practice

GEOTECHNICAL ENGINEERING STATE OF THE ART AND PRACTICE 135

Geotechnical Engineering State of the Art and Practice

136 GEOTECHNICAL ENGINEERING STATE OF THE ART AND PRACTICE

Geotechnical Engineering State of the Art and Practice

GEOTECHNICAL ENGINEERING STATE OF THE ART AND PRACTICE 137

Geotechnical Engineering State of the Art and Practice

138 GEOTECHNICAL ENGINEERING STATE OF THE ART AND PRACTICE

Documents that can be considered:

Geotechnical Engineering State of the Art and Practice

GEOTECHNICAL ENGINEERING STATE OF THE ART AND PRACTICE 139

Fig. 13 Flow chart for slope stability assessment

Evidences of geological processes such as landslides or erosion at the site and

Geotechnical Engineering State of the Art and Practice

140 GEOTECHNICAL ENGINEERING STATE OF THE ART AND PRACTICE

Geotechnical Engineering State of the Art and Practice

GEOTECHNICAL ENGINEERING STATE OF THE ART AND PRACTICE 141

Geotechnical Engineering State of the Art and Practice

142 GEOTECHNICAL ENGINEERING STATE OF THE ART AND PRACTICE