Professional Documents
Culture Documents
Municipal Solid Waste Slope Stability Modeling-A Probabilistic Approach
Municipal Solid Waste Slope Stability Modeling-A Probabilistic Approach
A Probabilistic Approach
Ali Jahanfar 1; Bahram Gharabaghi 2; Edward A. McBean 3; and Brajesh K. Dubey 4
Abstract: The geotechnical properties (e.g., density, friction angle, and cohesion) of waste material in landfills and dumpsites vary
Downloaded from ascelibrary.org by Florida Inst of Technology on 04/25/23. Copyright ASCE. For personal use only; all rights reserved.
significantly, depending on the waste composition and compaction practices. This variability results in a major source of uncertainty in
slope stability analysis for municipal solid waste (MSW) disposal sites. This study presents a novel probabilistic approach for slope stability
analysis by backcalculation, assigning probability distributions to waste material properties. Using backanalysis of historical, well-
documented failed waste slopes, probability distributions for the density, friction angle, and cohesion of MSW are assigned based on
the disposal method (landfill or dumpsite) and waste material composition (fibrous or food waste materials). MSW with (1) major food
waste contents has lower shear strength (friction angle = 22.2° and cohesion = 19.2 kPa) in comparison with MSW with major fibrous
waste contents (friction angle = 34° and cohesion = 14 kPa) and (2) compaction practices in landfills compared to dumpsites result in 20%
higher shear-strength properties for the waste material. The application of this new approach and its advantages is demonstrated/validated
using four case studies (Rumpke, Durban, Istanbul, and Sarajevo). DOI: 10.1061/(ASCE)GT.1943-5606.0001704. © 2017 American
Society of Civil Engineers.
Author keywords: Landfill failure; Probabilistic stability analysis; Monte Carlo; Waste friction angle; Waste cohesion; Waste unit weight.
Introduction the major reason for the low waste density in Payatas’s failure
was waste composition. He explained that scavenging of the
Landfill/dumpsite slope stability is a function of different factors, waste materials in the Payatas landfill had changed the waste
including pore water pressure, geometry, waste composition, and composition; high-strength materials such as wood, metal, card-
compaction (Stark and Huvaj-Sarihan 2009). Each of these factors, board, and intact bottles were scavenged by waste pickers, leav-
or a combination of these factors, may lead to the slope failure ing the waste with only organic and light plastic portions.
mechanism reaching critical conditions. • Waste compaction: Blight (2008) stated that lack of compac-
• Pore water pressure: For example, ten days of heavy rainfall tion can lead to landfill slope failure because, in the absence
in Payatas landfill and extra leachate injection in the Bogota of effective compaction of the waste in uncontrolled dumpsites
bioreactor landfill affected the pore water pressure and contrib- (e.g., Istanbul and Sarajevo dumpsites) or poor and inconsistent
uted to landfill slope failure (Blight 2008; Merry et al. 2005). compaction in landfills (e.g., Bogota and Durban landfills),
• Geometry: Height and side-slope are the main driving param- there is a relatively low–shear strength and cohesionless waste
eters in landfills that influence slope stability (Omari 2012; mass.
Babu et al. 2014). For example, the Rumpke landfill failure in Given the above, and the apparent variability of municipal solid
1996 happened when the landfill height exceeded the tolerable waste (MSW) geotechnical characteristics, challenges exist in se-
elevation by 13–15 m (Eid et al. 2000a). Also, landfill slope in lecting appropriate values for use in landfill design (Fowmes et al.
the Payatas dumpsite was estimated as 1.5 H:1 V prior to failure, 2006; Gharabaghi et al. 2008; Ering and Babu 2016; De Stefano
whereas a more typical value was 3 H:1 V (Merry et al. 2005). et al. 2016; Jahanfar et al. 2017). As a result, there is merit in using
• Waste composition: One might question why a designer uses a probabilistic approach to develop the distributions. The backcal-
typical friction angle values pertinent for U.S. landfills in the culation of landfill slope stability–related geotechnical characteris-
design of a landfill when dealing with sites in Asia where much tics is used as an example (Babu et al. 2014; Fowmes et al. 2006;
larger food percentages of waste contents exist (Dixon et al. Bray and Zekkos 2008; Jianguo et al. 2010; Onisiphorou et al.
2008; Moon and Ku 2016). Koelsch (2007) emphasized that 2011; Ramaiah et al. 2016; Sabatini et al. 2002; Srivastava and
Reddy 2012).
1
Ph.D. Candidate, Dept. of Civil Engineering, Univ. of Toronto, Probabilistic approaches have been used in landfill slope stabil-
35 St. George St., Toronto, ON, Canada M5S 1A4 (corresponding author). ity studies since 2002. In a recent study, Babu et al. (2014) used
E-mail: ali.jahanfar@mail.utoronto.ca numerical analysis and Monte Carlo simulation methods to assess
2
Professor, School of Engineering, Univ. of Guelph, 50 Stone Rd. East, the typical landfill slope stability within a probabilistic framework,
35 St. George St., Guelph, ON, Canada M5S 1A4. considering spatial variations of geotechnical properties. Compar-
3
Professor, School of Engineering, Univ. of Guelph, 50 Stone Rd. East, ing the results with a traditional factor of safety (FoS) value, the
35 St. George St., Guelph, ON, Canada M5S 1A4. mean value of the FoS distribution decreased with an increase in
4
Professor, Dept. of Civil Engineering, Indian Institute of Technology
the coefficient of variation. Srivasatava and Reddy (2012) used
Kharagpur, Kharagpur, West Bengal 721302, India.
Note. This manuscript was submitted on April 13, 2016; approved on
both deterministic and probabilistic methods of slope stability as-
January 12, 2017; published online on March 31, 2017. Discussion period sessment to calculate the FoS and reliability of typical landfills,
open until August 31, 2017; separate discussions must be submitted for respectively. A high range of FoS indicated safe slopes under
individual papers. This paper is part of the Journal of Geotechnical widely ranging shear-strength properties of a MSW while low reli-
and Geoenvironmental Engineering, © ASCE, ISSN 1090-0241. ability indices indicated safety concerns for a similar slope structure
To date, researchers have used both laboratory and in situ tests (Bareither et al. 2012). Hossain and Haque (2009) published one
to estimate unit weight, friction angle, and cohesion for MSW. of the few studies that has examined the effect of waste decompo-
Consequently, an extensive database of geotechnical parameters of sition on the shear strength of a laboratory-scale bioreactor landfill.
MSW is available to practitioners, although it is difficult to use and They observed considerable decreases in friction angle (28%) and
interpret this database in practice because of the heterogeneity of cohesion (47%) during the decomposition processes of MSW.
MSW. As a result, backanalysis of case studies of landfill slope However, bioreactor landfills and the effect of decomposition on
failures has been described herein as a reliable way of obtaining MSW shear strength are excluded from the assessment herein be-
representative geotechnical values for MSW (Singh et al. 2009; cause of the complexity and brevity of background experiments.
Ering and Babu 2016). Additional experimental and numerical studies on bioreactor land-
The preceding paragraphs demonstrate that there is substantial fills are required.
merit in using probabilistic methods for the analysis of landfill Dumping of waste still represents the primary waste disposal
slope stability. In response, a novel probabilistic approach for slope method in developing countries, because dumpsites entail a lower
stability analysis of landfills and dumpsites is described. Using operational cost in comparison with landfills (Munawar and Fellner
backanalysis of historical, well-documented failed waste slopes, 2013). With no daily cover soil, attaining sufficient compaction
an approach assigning probability distributions to density, friction and leachate drainage pose the greatest threats to structural stability
angle, and cohesion of MSW based on the disposal method (landfill of these types of disposal sites. Without an effective leachate col-
or dumpsite) and major waste contents (fibrous or food waste ma- lection system, the water content at dumpsites is potentially high,
terials) is provided. The application of the novel probabilistic slope especially in wet climate conditions, so that precipitation penetrates
stability analysis method and its advantages are demonstrated in throughout the MSW in the absence of daily cover soil and end
four case studies (Rumpke, Durban, Istanbul, and Sarajevo). cap layer.
Waste Composition
Waste Classification
In addition to operational aspects, knowledge of waste composition
Estimates of MSW shear strength and unit weight values are highly is usually one of the most valuable factors to estimate MSW proper-
uncertain (Bray and Zekkos 2009; Chen et al. 2010; Bray and ties (Bareither et al. 2012; Cho et al. 2011; Athanasopoulos et al.
Zekkos 2008). Uncertainties related to shear strength and unit 2008). The contribution of this factor in the MSW shear strength
weight depend on waste composition and disposal operation fac- data has been verified (Bray and Zekkos 2009; Zekkos et al. 2010),
tors (Bareither et al. 2012; Zekkos et al. 2006). Table 1 illustrates so it is considered in the above classification (Table 1). Wastes con-
a typical waste disposal classification based on the disposal method tain different categories of materials, e.g., fibrous material varies
(landfill or dumpsite) and major waste contents (fibrous or food from low-degradable (e.g., paper, cardboard, texture, wood, gravel)
waste materials). to nondegradable (e.g., metal, plastics, glass) materials (Sharma and
McBean 2007). These materials are identified as the primary cause
Disposal Method (Waste Compaction) of the large shear strength of MSW as a result of interlocking
(Zekkos et al. 2010; Wong 2009; Cho et al. 2011; Bareither et al.
Globally, landfilling is the primary engineered method for MSW 2012). Food waste (e.g., cooked food, uncooked vegetables and
disposal. There are two types of landfills: conventional landfills fruits) contents are biodegradable. The approximate food waste
and bioreactor landfills. Conventional landfills contain a bottom content of U.S. MSW is approximately 12.5%, while refuse in
liner, leachate and gas collection system, and daily soil cover to China has been indicated as up to 73% food waste (Cho et al. 2011).
avoid leachate and contamination release into the environment. Hence, large variations in food waste materials within MSW can be
Compaction effort is the main factor that controls the short- an important factor to explain the variation in MSW shear strength
term density and resulting shear strength of waste in landfills and disposal slope stability. In addition to municipal waste materi-
als, landfills usually contain up to 30% daily cover soils (nondegrad-
able constituent), which are mixed with MSW layers during the
Table 1. Waste Disposal Classification Based on Disposal Method and operation time (Machado et al. 2002). This layer has a reinforcement
Major Waste Contents effect on the slope stability of landfills (Hossain and Haque 2009).
Waste class Disposal method Major waste contents
Class I Landfill Fibrous material Waste Unit Weight
Class II Landfill Food waste
Class III Dumpsite Fibrous material Zekkos et al. (2006) studied the MSW unit weight based on
Class IV Dumpsite Food waste
waste compaction and soil contents. From literature data,
100
30
50
40 0
50 100 150 200 250 300
50 Normal Stress (kPa)
Class I
60 Class II
Downloaded from ascelibrary.org by Florida Inst of Technology on 04/25/23. Copyright ASCE. For personal use only; all rights reserved.
Low compaction and soil content (Class III & IV) Class I (Upper & Lower bound)
Class II (upper & Lower bound)
Typical compaction and soil content (Class I & II) Zekkos et. al., (2012)
Cho et. Al., 2011 (Major Food-waste)
High moisture waste or soil content
Cho et. Al., 2011 (Major Fibrous-Material)
Stark et al. (2009) recommendation
Fig. 1. Unit weight field-fitted hyperbolic curves for different classes
of waste (adapted from Zekkos et al. 2006) Fig. 2. MSW shear strength versus normal stress for Class I and
Class II
Cohesion (kPa)
20 Class III
spatial and temporal variability and waste heterogeneity, increasing
Class IV
statistical uncertainties. Furthermore, unit weight, friction angle,
and cohesion are assumed to be statistically independent. Table 2 10
summarizes the mean and standard deviation of unit weight, fric-
tion angle, and cohesion for different classes of waste mass disposal
0
methods. 0 10 20 30 40 50
In terms of friction angle and cohesion for Classes I and II, the Friction Angle
most likely strength envelope is calculated using linear regression
(Fig. 2). The mean friction angle and cohesion for Classes III and Fig. 3. Uncertainty range of friction angle and cohesion of Classes I
IV have been assumed to be 20% less than those for Classes I and to IV
Downloaded from ascelibrary.org by Florida Inst of Technology on 04/25/23. Copyright ASCE. For personal use only; all rights reserved.
ðHCV − LCVÞ as linear models and hence the shear strength parameters would
SD ¼ ð1Þ primarily depend on the composition and compaction of MSW.
6
The lower and upper bounds in Fig. 2 for each class can be used Case Studies
to estimate the highest conceivable value (HCV) and lowest con-
ceivable value (LCV) of the strength envelope for calculation of the Case study failures for two dumpsites (Sarajevo 1977 and Istanbul
SD of the waste shear strength. Using the three-sigma rule, SD can 1993) and two landfills (Rumpke 1997 and Durban 1997) are de-
be estimated because 99.7% of all normally distributed values fall scribed herein. Each failure case study is assigned to one class of
within three SDs plus or minus a mean value (Duncan 2000). Stark MSW based on the effect of compaction and composition on geo-
and Huvaj-Sarihan (2009) recommended this method to estimate technical factors of MSW (Table 3).
the FoS probabilistic distribution in landfill slope stability instead In addition to geotechnical factors, the water pressure along the
of simply reporting a value of FoS. slip surface affects slope stability. El-Ramly et al. (2002) described
In Fig. 3, the range of the friction angle and cohesion for the pore pressure ratio (Ru ), the ratio of water pressure to total nor-
Classes I to IV are compared to each other. mal stress, as a useful parameter to characterize the water pressure
Regarding Classes II and IV, high food waste contents decrease of the slip surface in complex situations. Assuming a normal dis-
the friction angle to the lower level than Classes I and III with lower tribution for Ru, the mean value for each failed slope is provided,
food waste contents. Thus, a friction angle of MSW with minor based on knowledge/experience with the groundwater regime gov-
food waste contents (Classes I and III) falls within a typical range erned by literature studies (Table 3). The SD of 0.05 is used to
of 21.2–40°, while the friction angle decreases to the range of 14.7– account for the lack of site-specific measurements.
24° for MSW with major food waste contents (Classes II and IV). Fig. 4 illustrates cross-section geometries for all case studies.
Also, for minimal compaction and daily soil cover, the best esti- Lateral translational failure is the most probable type of failure for
mates of friction angle and cohesion for Classes III and IV are dumpsites and landfills. Thus, fully specified block shapes have
reduced by 20% from those stated for Classes I and II. been used as the slip surface in all case studies.
Stark et al. (2009) presented strain and stress dependency of the
Mohr-Coulomb shear failure envelope of MSW. However, large
Rumpke Landfill Failure (1996)
shear displacements that are required to mobilize the peak shear
strength of MSW can lead to undesirable displacement incompat- The Rumpke landfill, also known as Rumpke Mountain, in
ibility between MSW and the underlying geosynthetic material. Cincinnati, Ohio, is one of the largest landfills in the United States.
Therefore, conservative shear-strength parameter values are used In March 1996, the largest slope failure in the United States hap-
for the design of stable slopes by setting stringent failure criteria pened at this landfill a few days following the 45-m extra excava-
limits for the maximum allowable MSW shear displacements tion in front of the landfill toe (Eid et al. 2000a). In addition, the site
(<25 mm) and maximum axial strain (<10%). For these lower was overfilled by 13–15 m at the time of the failure.
ranges of displacements (and small axial strains <10%), the Mohr- Eid et al. (2000a) indicated that the failure surface developed
Coulomb shear strength envelope of MSW can be approximated through a nearly vertical cut in the MSW and a horizontal cut in the
saturated brown native soil underlying the waste [Fig. 4(a)]. With a
typical U.S. waste type (low food waste contents) and the presence
Table 2. Average and Standard Deviation Unit Weight, Friction Angle, and of waste compaction and daily soil cover, this landfill is approx-
Cohesion Discussed for Different Classes of Waste Mass Disposal Methods imately categorized as Class I. The probabilistic distribution of geo-
technical characteristics of the MSW are extracted from Class I in
Shear strength
Design Unit weight Friction Cohesion
class (kN=m3 ) angle (degrees) (kPa)
Table 3. Failed Slopes, Assigned Classes and Pore Water Pressures
Class I 10–14 (1) 34 (2) 14 (1)
Failed slope Assigned class Ru mean value References
Class II 10–14 (1) 22.2 (1) 19.2 (2)
Class III 5–12.5 (1) 27.2 (2) 11.2 (1) Rumpke (1996) Class I 0.3 Eid et al. (2000a)
Class IV 5–12.5 (1) 17.7 (1) 15.4 (2) Durban (1997) Class II 0.67 Blight (2008)
Sarajevo (1977) Class III 0.42
Note: Values in the table are the mean and in the brackets (standard
Istanbul (1993) Class IV 0.14
deviation).
Fig. 4. (a) Rumpke and (b) Durban landfill cross-sectional geometries before failure; (c) Sarajevo and (d) Istanbul dumpsite cross-section geometries
before failure
Table 2, while the geotechnical characteristics of subsoil are as- To simulate this situation, Blight (2008) used an Ru value of 0.67
sumed constant. Eid et al. (2000a) suggested a unit weight, friction for the backanalysis of Durban landfill slope stability. This value is
angle, and cohesion of 19 kN=m3 , 17°, and 0 kPa for subsoil, assumed to be an average of the initial distribution of Ru with a SD
respectively. The mean and SD of the Ru prior distribution are of 0.05 for backcalculation.
assumed to be 0.3 and 0.05.
The interconnection of the fibrous materials contributes to the Sarajevo Dumpsite Failure (1977)
high shear strength of the MSW in this case history. In addition, the
probable failure surface likely extended to deeper surfaces because The Sarajevo dumpsite failure was the first recorded waste disposal
of initial deep cracks at the top of the slope and a potential block failure in the technical literature. In December 1977, approximately
form slide. Thus, stability evaluation depends on the shear strength 200,000 m3 of waste slid down the dumpsite slope and traveled
of the foundation soil rather than that of the MSW (Eid et al. more than 1 km from the site. This dumpsite contained 74% fibrous
2000b). This case history is a great example to show that although materials, including 45% paper and cardboard. No compaction
waste compaction and composition are contributory factors, foun- value was reported for this dumpsite. Blight (2008) indicated that
dation design can also be a critical factor in landfill/dumpsite slope a translational wedge failure took place at this dumpsite, which was
stability. potentially the result of winter rain infiltration into the uncom-
pacted and uncovered waste [Fig. 4(c)].
Durban Landfill Failure (1997) The Sarajevo dumpsite failure is an example of Class III waste
disposal. Hence, input data for a slip surface through MSW and
The Durban landfill (also called Bulbul landfill), in Durban, South along the interface of the MSW and ground surface are adopted
Africa, was designed to be filled with a specific codisposal ratio of from strength properties of MSW from Class III in Table 2. The
liquid and dry waste. Dry waste contained 31% food waste and friction angle and cohesion in Class III are 20% lower than those
liquid waste was industrial sludge, indicating that a major waste of Class I, as a result of the lack of compaction and soil cover. Since
content of this landfill was nonfibrous materials. To increase slope a high pore water pressure may have caused this failure, an Ru value
stability, berms were constructed across the toe of the landfill for with a mean and standard deviation of 0.42 and 0.05, respectively,
each phase of waste disposal. are assumed as inputs for backcalculation.
In September 1997, the landfill failed, and the waste moved
on to an area prepared for a future phase of the landfill [Fig. 4(b)].
Istanbul Dumpsite Failure (1993)
A translational failure occurred within a face of a previous land-
filling phase on the hillside and within a weak boundary cover layer About 30 km from Istanbul city, the Istanbul dumpsite is located
(clay liner) and underlying soil in the foothill of the slip surface, near the edge of a steep hillside [Fig. 4(d)]. Heavy rain and exces-
respectively. The MSW in this landfill can be classified as Class II sive infiltration, together with a suddenly applied disturbing force,
because of the provided compaction and low fibrous materials. The triggered the failure, causing 39 deaths from the waste’s movement
input data for probabilistic distributions of MSW geotechnical and destroying 11 informal brick-built houses. Blight (2008) re-
characteristics are collected from Class II in Table 2. The geotech- ported that 72% of the waste contents were food waste in this
nical characteristics of clay liner are assumed to be constant (unit dumpsite. The waste materials were placed without any protection
weight ¼ 18 kN=m3 , friction angle = 12°, and cohesion = 10 kPa). at the base, and there was no compaction or daily cover of the waste
A high pore water pressure in the waste body as a result of the at the site.
omitted drainage provisions of the design and greater volumes of The Istanbul dumpsite failure can be classified as Class IV waste
liquid with the same mass of dry waste was the likely reason for disposal because of the lack of compaction and high range of food
the failure (Blight 2008). Considering the evidence, the berms and waste. Class IV has a 20% lower friction angle and cohesion than
liners provided a huge barrier around the saturated waste materials. Class II. The strength properties of MSW from Class IVare used for
the slip surface through MSW, and along the interface of the MSW parameters in slope stability analysis and this procedure, with
and ground surface, as input data for the Istanbul dumpsite back- details as outlined in the Appendix, was followed.
analysis (Table 2). As discussed above, the parameters to be backcalculated for
Blight (2008) obtained Ru ¼ 0.14 for this failure, while Stark waste disposal slopes can be denoted by the vector θ ¼ fCW ; ⊘W ;
and Huvaj-Sarihan (2009) calculated a value of 0.41. The prior γ W ; Ru gT . Assuming μθ as the mean and Cθ as a covariance ma-
mean and SD of Ru are assumed to be 0.14 and 0.05 as input data trix of θ prior distribution, the objective of probabilistic backanal-
for backcalculation. ysis is then to improve the probabilistic distribution of θ to μðθjdÞ
and CðθjdÞ
∂gðθÞ 2 H ¼ f0.0188; 0.0168; −0.0636; −1.2993g
varðFoSÞ ≈ × varðCw Þ
∂cw θ ¼ μθ
The sensitivity of FoS to γ W and Ru are negative, meaning that
∂gðθÞ 2 FoS decreases with the growth of either γ W or Ru. In this case, FoS
þ × varð⊘w Þ
∂⊘w θ ¼ μθ is respectively sensitive to Ru , γ W , CW , and ⊘W . Using mean and
SD values for CW , ⊘W , γ W , and Ru and the Durban landfill geom-
∂gðθÞ 2 etry as inputs to the Slope/W model, a probabilistic distribution
þ × varðγ w Þ
∂γ w θ ¼ μθ of FoS is obtained. Fig. 6 illustrates the calculation process in an
Excel sheet.
∂gðθÞ 2
þ × varðRu Þ þ σ2ε ð5Þ The mean of the FoS distribution based on the prior distribution
∂Ru θ ¼ μθ is 1.1, with a probability of failure of 0.4 (Step 1). As the FoS mean
value is not equal to unity, Eqs. (2) and (3) are used to backcalculate
the CW , ⊘W , γ W , and Ru distributions. The FoS after one backcal-
culation is 1.00, with a probability of failure of 0.51 (Step 2). Thus,
Example the backcalculated CW , ⊘W , γ W , and Ru distributions are qualified
The backcalculation of cohesion, friction angle, unit weight, and as posterior distributions (Step 3).
pore pressure ratio for the Durban landfill is described below.
The prior distributions of CW , ⊘W , γ W , and Ru can be described
with a multivariate normal distribution with a mean of μθ ¼ Result of Backanalysis (Posterior Data)
f19.2; 22.2; 13; 0.67gT and the following covariance matrix (after Two dumpsite failures (Sarajevo 1977 and Istanbul 1993) and two
Tables 2 and 3): landfill failures (Durban 1997 and Rumpke 1996) are described
2 3 2 3 herein. The geotechnical characteristics of the MSW in each failure
SD2Cw 0 0 0 22 0 0 0 have been addressed based on corresponding classifications in
6 7 6 7
6 0 SD2⊘w 0 0 7 12 Table 1. The initial distributions of unit weight, friction angle, and
6 7 6 0 0 0 7
Cθ ¼ 6 7¼6
6
7
7 cohesion of the waste materials (Table 2) were utilized as inputs
6 0 0 SD2γw 7
0 5 40 0 12 0 5
4 on the uncertainty in the backanalysis. Using Eqs. (1)–(3), the pos-
0 0 0 SD2Ru 0 0 0 0.052 terior distributions of CW , ⊘W , γ W , and Ru for each slope failure
are listed in Table 4.
The units for CW , ⊘W , γ W , are kPa, degree, and kN=m3 ,
respectively. Information Allocation in Backanalysis
The measured slope representing the sensitivity of FoS to CW ,
⊘W , γ W , and Ru are described with a row vector as follows The Sarajevo and Istanbul dumpsites, as well as Rumpke and
[Eq. (4)]: Durban landfill failures, were backanalyzed in terms of slope
a
As MSW unit weight changes with depth, means of unit weight distributions in −25m are listed in here.
Sensitivity Analysis
Summary and Conclusion
Using the probabilistic backanalysis method, prior distributions of
all input parameters were modified by slope failure events simul- This study is the first of its kind that presents a framework for a
taneously and posterior distributions were obtained. A sensitivity probabilistic slope stability analysis for landfills and dumpsites.
analysis is required to realize the degree of modification achieved This novel approach takes advantage of the information about the
for each parameter by the backanalyses. waste material compaction and composition to assign probability
and Sarajevo) are comparable with sandy soils, while MSW with Manage., 31(1), 26–32.
major contents of food waste materials (e.g., Durban and Istanbul) Christian, J. T., Ladd, C. C., and Baecher, G. B. (1994). “Reliability applied
to slope stability analysis.” J. Geotech. Eng., 10.1061/(ASCE)0733
are cohesive.
-9410(1994)120:12(2180), 2180–2207.
Considering waste compaction and composition as two con-
De Stefano, M., Gharabaghi, B., Clemmer, R., and Jahanfar, M. (2016).
tributory factors in landfill/dumpsite stability, the following conclu- “Berm design to reduce risks of catastrophic slope failures at solid waste
sions are drawn for MSW unit weight, friction angle, and cohesion disposal sites.” Waste Manage. Res., 34(11), 1117–1125.
in different scenarios: Dixon, N., Langer, U., and Gotteland, P. (2008). “Classification and
1. The backcalculated solid waste material unit weight distribu- mechanical behavior relationships for municipal solid waste: Study us-
tions primarily depend on the composition and compaction ing synthetic wastes.” J. Geotech. Geoenviron. Eng., 10.1061/(ASCE)
practices and change nonlinearly with depth. A lack of compac- 1090-0241(2008)134:1(79), 79–90.
tion is best described by a unit weight ranging from 5 kN=m3 at Duncan, J. M. (2000). “Factors of safety and reliability in geotechnical en-
the high end to 12.5 kN=m3 at the low end, with a standard de- gineering.” J. Geotech. Geoenviron. Eng., 10.1061/(ASCE)1090-0241
viation of 1 kN=m3 . Compaction effort raises the unit weight to (2000)126:4(307), 307–316.
a typical value between 10 and 14 kN=m3 at the top and bottom, Eid, B. H. T., Stark, T. D., Evans, W. D., and Sherry, P. E. (2000). “Munici-
pal solid waste slope failure. I: Waste and foundation soil properties.”
with a standard deviation of 1 kN=m3 , respectively.
J. Geotech. Geoenviron. Eng., 10.1061/(ASCE)1090-0241(2000)126:
2. Friction angle and cohesion of fibrous waste buried in engi- 5(397), 397–407.
neered landfills (which include waste compaction and daily soil Eid, H. T., Stark, T. D., Evans, W. D., and Sherry, P. E. (2000). “Municipal
cover) are best fitted by normal distributions with mean values solid waste slope failure. II: Stability analyses.” J. Geotech. Geoen-
of 34° and 14 kPa and standard deviations of 2° and 1 kPa, viron. Eng., 10.1061/(ASCE)1090-0241(2000)126:5(397), 408–419.
respectively. El-Ramly, H., Morgenstern, N. R., and Cruden, D. M. (2002). “Probabi-
3. The friction angle and cohesion of solid waste material with listic slope stability analysis for practice.” Can. Geotech. J., 39(3),
high food waste content disposed in engineered landfills can be 665–683.
described with normal statistical distributions with mean values Ering, P., and Babu, G. S. (2016). “Probabilistic back analysis of rainfall
of 22.2° and 19.2 kPa and standard deviations of 1° and 2 kPa, induced landslide: A case study of Malin landslide, India.” Eng. Geol.,
respectively. 208, 154–164.
4. In dumpsite cases (which do not have waste compaction Fowmes, G. J., Jones, D. R. V., and Dixon, N. (2006). “Interface shear
strength variability and its use in reliability-based landfill stability
and daily soil cover), the representative average shear strength
analysis.” Geosynthetics Int., 13(1), 1–14.
parameters of MSW decrease by 20% from the shear strength Gharabaghi, B., Singh, M. K., Inkratas, C., Fleming, I. R., and McBean,
recommended for landfill cases. E. A. (2008). “Comparison of slope stability in two Brazilian municipal
landfills.” Waste Manage., 28(9), 1509–1517.
Hanson, J. L., Yesiller, N., Von Stockhausen, S. a., and Wong, W. W.
Acknowledgments (2010). “Compaction characteristics of municipal solid waste.”
J. Geotech. Geoenviron. Eng., 10.1061/(ASCE)GT.1943-5606
This research was made possible by funding from the Natural .0000324, 1095–1102.
Sciences and Engineering Research Council of Canada. Hossain, M. S., and Haque, M. a. (2009). “The effects of daily cover soils
on shear strength of municipal solid waste in bioreactor landfills.”
Waste Manage., 29(5), 1568–1576.
References Jahanfar, A., Amirmojahedi, M., Gharabaghi, B., Dubey, B., McBean, E.,
and Kumar, D. (2017). “A novel risk assessment method for landfill
Athanasopoulos, G., Grizi, A., Zekkos, D., Founta, P., and Zisimatou, E. slope failure: Case study application for Bhalswa Dumpsite, India.”
(2008). “Municipal solid waste as a reinforced soil: Investigation using Waste Manage. Res., 35(3), 220–227.
synthetic waste.” ASCE-Geoinstitute Geocongress, ASCE, Reston, VA, Jianguo, J., Yong, Y., Shihui, Y., Bin, Y., and Chang, Z. (2010). “Effects of
168–175. leachate accumulation on landfill stability in humid regions of China.”
Babu, G. L. S., Asce, M., Reddy, K. R., Asce, F., Srivastava, A., and Asce, Waste Manage., 30(5), 848–855.
A. M. (2014). “Influence of spatially variable geotechnical properties Koelsch, F. (2007). “Stability problems of landfills—The Payatas
of MSW on stability of landfill slopes.” J. Hazard. Toxic Radioactive landslide.” Proc., 17th Int. Conf. on Solid Waste Technology and Man-
Waste, 10.1061/(ASCE)HZ.2153-5515.0000177, 27–37. agement, Widener Univ. School of Engineering, Chester, PA, 1–7.
Bareither, C. A., Benson, C. H., and Edil, T. B. (2012). “Effects of waste Machado, S., Carvalho, M., and Vilar, O. (2002). “Constitutive model for
composition and decomposition on the shear strength of municipal solid municipal solid waste.” J. Geotech. Geoenviron. Eng., 10.1061/(ASCE)
waste.” J. Geotech. Geoenviron. Eng., 10.1061/(ASCE)GT.1943-5606 1090-0241(2002)128:11(940), 940–951.
.0000702, 1161–1174. McBean, E. a., Syed-Ritchie, S., and Rovers, F. a. (2007). “Performance
Blight, G. (2008). “Slope failures in municipal solid waste dumps and results from the Tucuman solid waste bioreactor.” Waste Manage.,
landfills: A review.” Waste Manage. Res., 26(5), 448–463. 27(12), 1783–1791.