522492

research-article2014

WMR0010.1177/0734242X14522492Waste Management and ResearchGiri and Reddy

Original article

Slope stability of bioreactor landfills

during leachate injection: Effects of
heterogeneous and anisotropic
municipal solid waste conditions

Waste Management & Research

2014, Vol. 32(3) 186197
The Author(s) 2014
Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0734242X14522492
wmr.sagepub.com

Rajiv K Giri and Krishna R Reddy

Abstract
In bioreactor landfills, leachate recirculation can significantly affect the stability of landfill slope due to generation and
distribution of excessive pore fluid pressures near side slope. The current design and operation of leachate recirculation
systems do not consider the effects of heterogeneous and anisotropic nature of municipal solid waste (MSW) and the increased
pore gas pressures in landfilled waste caused due to leachate recirculation on the physical stability of landfill slope. In this
study, a numerical two-phase flow model (landfill leachate and gas as immiscible phases) was used to investigate the effects
of heterogeneous and anisotropic nature of MSW on moisture distribution and pore-water and capillary pressures and their
resulting impacts on the stability of a simplified bioreactor landfill during leachate recirculation using horizontal trench
system. The unsaturated hydraulic properties of MSW were considered based on the van Genuchten model. The strength
reduction technique was used for slope stability analyses as it takes into account of the transient and spatially varying porewater and gas pressures. It was concluded that heterogeneous and anisotropic MSW with varied unit weight and saturated
hydraulic conductivity significantly influenced the moisture distribution and generation and distribution of pore fluid
pressures in landfill and considerably reduced the stability of bioreactor landfill slope. It is recommended that heterogeneous
and anisotropic MSW must be considered as it provides a more reliable approach for the design and leachate operations in
bioreactor landfills.
Keywords
Bioreactor landfill, capillary pressure, leachate recirculation, moisture distribution, municipal solid waste, pore-water pressure,
slope stability

Introduction
In recent years, bioreactor landfills have emerged as a successful
means for the safe disposal of municipal solid waste (MSW). In
bioreactor landfills, the collected leachate is recirculated into the
MSW, in addition of supplementary liquids, to increase the
moisture and result in fast biodegradation of MSW due to
enhanced microbial activity (Barlaz etal., 1992; Chugh etal.,
1998; Reinhart etal., 2002). Horizontal trenches (HTs), vertical
wells, and/or drainage blankets are employed as leachate recirculation systems to recirculate leachate in bioreactor landfills.
Constant injection pressure for a specified time period is needed
to add moisture to the landfill (Xu etal., 2012). However, high
injection pressures in leachate recirculation systems near the
side slopes can generate excess pore fluid (i.e. water and gas)
pressures and reduce the shear strength of the MSW due to
decreased effective stress, which may endanger the stability of
the landfill slope.
Engineered landfill design and operation should consist
of a careful assessment of landfill slope stability. Often,

these landfills are constructed near highly populated areas,

which further increases the risk associated with landfill slope
failure. Slope stability analyses of conventional landfills
based on the geotechnical properties of MSW and underlying
soils (i.e. unit weight, shear strength) have been reported in
the literature (Eid etal., 2000; Mitchell etal., 1990;
Gharabaghi etal., 2008; Zhan etal., 2008). In recent years,
analyses of landfill slope stability during leachate operations
have been getting more attention. For example, Koerner and
Soong (2000) studied numerous landfill slope failures to
analyse the developed pore-water pressures due to leachate
University of Illinois at Chicago, Chicago, IL, USA
Corresponding author:
Krishna R Reddy, Department of Civil and Materials Engineering,
University of Illinois at Chicago, 842 West Taylor Street, Chicago, IL
60607, USA.
Email: kreddy@uic.edu

187

Giri and Reddy

injection that resulted in slope failures. Kavazanjian and
Merry (2005) reported the Payatas landfill failure due to elevated levels of leachate that led to excessive pore-water
pressure in the landfilled MSW. Xu etal. (2012) carried out
a single-phase flow modelling to determine the effects of
pressurized leachate injection on the stability of a simplified
bioreactor landfill slope.
However, these studies only considered the effects of porewater pressure and neglected the pore gas (air) pressure. Also, the
MSW was considered to be homogeneous, which does not represent the true nature of landfilled waste, as the MSW is found to
be heterogeneous in nature with hydraulic conductivity, unit
weight, and shear strength parameters varying with depth due to
overburden stress and degree of decomposition (DOD) (Reddy
etal., 2011). Using a numerical two-phase flow model, Reddy
etal. (2013) showed that gas pressures can significantly exceed
liquid pressure in a typical bioreactor landfill configuration during initial phases of leachate injection. Therefore, it is critical to
evaluate the impact on the stability of bioreactor landfill slope
particularly in response to moisture distribution and pore liquid
and coupled gas pressures generated due to leachate injection
using HTs for heterogeneous and anisotropic MSW (HTAW)
conditions.
In this study, a numerical two-phase flow modelling was used
to determine the effects of heterogeneous and anisotropic nature
of MSW under elevated injection pressure on the moisture distribution, generation and distribution of pore-water and capillary
pressures (i.e. difference in pore gas pressure and water pressure)
within the landfill, and, ultimately, the resulting impact on the
stability of bioreactor landfill slope. The two-phase flow model
validation has been presented elsewhere (Giri and Reddy, 2013)
based on previously published studies using a single-phase flow
model and stability analyses for a simplified bioreactor landfill
configuration incorporating homogeneous and anisotropic MSW
(HAW).

Methods
Numerical two-phase flow and slope
stability model
The pores of unsaturated MSW were assumed to be filled with
two immiscible fluids: namely the landfill leachate and landfill
gas. The two-phase flow model incorporated modelling the flow
of these two immiscible fluids (i.e. leachate considered as wetting fluid and landfill gas considered as nonwetting fluid). The
capillary pressure is a function of leachate degree of saturation
and can be represented using the model of van Genuchten (1980).
The flow of leachate and landfill gas was described by Darcys
law, whereas relative permeability of each fluid is based on leachate saturation by the empirical laws of van Genuchten function
(ITASCA Consulting Group, 2011). In the numerical two-phase
flow model, the governing equations of unsaturated MSW are
given by the linear momentum balance and the fluid mass balance laws and are represented as:

= d + n ( S L L + SG G )

(equation 1)

q L
S P S
n L L + L = i
t
K L t
xi

(equation 2)

q G
S P S
n G G + G = i
t
K G t
xi

(equation 3)

where n is porosity, SL is leachate (liquid) saturation, SG is gas

saturation, PL is pore liquid pressure, PG is pore gas pressure, L
and G are fluid densities, d is matrix dry density, KL and KG are
liquid and gas bulk modulus, respectively, and qiL and qiG are
flow rate of liquid and gas given by Darcys law.
The governing equations 13 were solved numerically with
the Fast Lagrangian Analysis of Continua (FLAC) program using
the finite difference method. The detailed mathematical formulations including governing equations related to the two-phase
flow model are explained elsewhere (ITASCA Consulting Group,
2011, Reddy etal., 2013).
Concurrently, slope stability analyses were performed
using FLAC, wherein the strength reduction technique was
adopted to compute factor of safety (FOS; Dawson etal.,
1999). The MohrCoulomb failure criterion was combined
with the strength reduction approach for stability analyses. In
this approach, the FOS calculation was performed by successively reducing the shear strength parameters (cohesion and
friction angle) of MSW until the slope reached on the verge
of failure. Further information on the modelling approach and
its successful application is presented elsewhere (Giri and
Reddy, 2013).

Landfill configurations
A two-dimensional bioreactor landfill, 175 m wide and 50 m
deep with a side slope of 3:1 (horizontal/vertical), was created in
FLAC to investigate the effects of HTAW under pressurized leachate addition. Figure 1a shows the landfill cell, known as base
scenario, wherein the leachate was injected through a HT with a
continual injection pressure of 49 kPa (i.e. equivalent to a 5-m
water column head). The landfill model was considered to be
completely filled with a homogeneous and anisotropic waste
(HAW) throughout its entire depth for the numerical two-phase
flow model validation (Figure 1a). The wet zone contour represented the extent of moisture surrounding the HT due to leachate
injection. The top boundary was extended to a width of 25 m
away from the side slope. A 0.3-m-thick leachate collection-andremoval system, consisting of free draining granular soil, was
assumed to be located at the bottom of the landfill. A HT (11 m)
was placed at an elevation of 30 m above the base of the leachate
collection-and-removal system and at a setback of 30 m from the
side slope. The landfill configuration was similar to that reported
by Xu etal. (2012), who used the single-phase flow model

188

Waste Management & Research 32(3)

Figure 1. (a) Simplified bioreactor landfill configuration with homogeneous anisotropic waste (HAW), depicting the base
scenario; (b) landfill configuration, depicting heterogeneous and anisotropic waste (HTAW).

SEEP/W and SLOPE/W, respectively, to evaluate pore-water

pressures and their impact on slope stability analysis. The conceptual model did not consider the effects of a landfill cover system because this study was mainly focused on pressurized
leachate injection and flow through landfilled waste when bioreactor landfill is in active state.
Temperature effects, mechanical compression, and infiltrations were not included in the model. In bioreactor landfills, various biochemical processes and waste degradation generate
significant amount of heat that result in long-term elevated waste
temperatures (Yeiller etal., 2005). The relatively high waste temperatures (~3045C) affect the process of landfilled waste
decomposition that further varies the waste physical (e.g. void
ratio, deformation) and geotechnical properties (e.g. unit weight,
shear strength) as well as hydraulic properties (e.g. moisture content, saturation, pore pressure, leachate infiltration). The effects of
elevated temperatures on the aforementioned waste properties are
assumed to be even more prominent in bioreactor landfills than
conventional dry landfills and should be considered for future
research studies. However, the present study did not account for
these coupled thermo-hydro-bio-mechanical interactions.
To investigate the effects of HTAW conditions, the 50-m-deep
landfill model was divided into 10 different layers, each layer

having a depth of 5 m. The unit weights, saturated hydraulic conductivities, and shear strength of MSW varied with depth and are
explained in the next section. Figure 1b depicts a landfill configuration taking HTAW scenarios into account.

Material properties
Limited data are available on shear strength of MSW and further
research is needed to accurately predict the variation in shear
strength properties of MSW with depth during leachate recirculation (Reddy etal., 2009a). Variation in unit weight of MSW with
depth was given using the relationship proposed by Zekkos etal.
(2006):

= i +

z
+z

(equation 4)

where is unit weight at depth z, and are modelling parameters for typical MSW, and i is near surface in-place unit weight.
The saturated hydraulic conductivity of MSW decreases with
depth due to the increase in normal stress caused by overlying
MSW and this can be expressed by the relationship proposed by
Reddy etal. (2009b):

189

Giri and Reddy

Table 1. Variations of heterogeneous and anisotropic MSW properties considered for model simulations.
Layer

Depth
(m)

HTAW-1
(kN
m3)

kv (cm
s1)

()

c (kPa)

(kN
m3)

kv (cm
s1)

()

c
(kPa)

DOD
(%)

(kN
m3)

kv (cm
s1)

()

c
(kPa)

05
510
1015
1520
2025
2530
3035
3540
4045
4550

15
15
15
15
15
15
15
15
15
15

1.9103
1.9104
3.9105
1.1105
4.2106
1.8106
8.9107
4.7107
2.7107
1.6107

35

15

12.6
13.5
14.1
14.6
14.9
15.1
15.3
15.4
15.6
15.7

2.4103
2.9104
5.9105
1.7105
5.9106
2.4106
1.1106
5.7107
3.1107
1.8107

35

15

0
10
20
30
40
50
60
70
80
90

14.9
17.2
17.9
18.4
18.7
19.0
19.2
19.4
19.5
19.7

1.9103
1.3104
2.1105
5.2106
1.7106
6.6107
3.0107
1.5107
7.8108
4.4108

35.0
33.9
32.9
31.8
30.8
29.7
28.7
27.6
26.6
25.5

15.0
16.6
18.1
19.7
21.2
22.8
24.3
25.9
27.4
29.0

10 (Top)
9
8
7
6
5
4
3
2
1 (Bottom)

HTAW-2

HTAW-3

, unit weight; kv, vertical hydraulic conductivity; , friction angle; c, cohesion; DOD, degree of decomposition.

Table 2. Unsaturated hydraulic MSW parameters based on

Breitmeyer and Benson (2011).
Parameter

Value
m3)

Unit weight (kN

Inverse of air-entry pressure (kPa1)
Saturated moisture content s
Residual moisture content r
van Genuchten steepness parameter n
van Genuchten m

'
kv = kv 1 +
pa
0

7.8
1.18
0.41
0.03
1.33
0.25

5.3

(equation 5)

where kv0 is initial saturated hydraulic conductivity at zero normal stress (102 cm s1), kv is saturated hydraulic conductivity
under effective overburden of , and pa is atmospheric pressure
MSW is heterogeneous and anisotropic in nature. Therefore,
to investigate this condition systematically, the bioreactor landfill
cell (Figure 1b) was modelled for three different HTAW conditions with varying geotechnical properties with depth, and the
results were compared with simplified HAW:
HAW: For this condition, the MSW properties were assumed
to be the same for entire depth of the landfill and were directly
adopted from Xu etal. (2012). Unit weight (), cohesion (c),
friction angle (), vertical saturated hydraulic conductivity
(kv), and anisotropy (a; kh/kv) were set at 15 kN m3, 15 kPa,
35, 105 cm s1, and 10 (Tchobanoglous et al., 1993).
HTAW-1: The unit weight, anisotropy, and shear strength of
MSW were taken to be exactly the same as that of HAW
across the landfill cell. However, the saturated hydraulic conductivity for each layer was varied and calculated using equation 5, depending on the estimated overburden stress at the
centre of each layer with the MSW unit weight of 15 kN m3.
Table 1 shows the MSW properties for HTAW-1.

HTAW-2: This represented a realistic heterogeneous nature

of MSW found in bioreactor landfills immediately after
placement, for which the unit weight and saturated hydraulic
conductivity of MSW were varied with depth due to overburden stress. The unit weight of the MSW at the mid depth
(25 m) of the landfill cell was exactly the same as that of
HAW (i.e. =15 kN m3) and unit weights for rest of the layers
were varied with depth using equation 4. Saturated hydraulic
conductivity of each waste layer decreased with depth and
was calculated using equation 5. Shear strength and anisotropy (value 10) of MSW were constant throughout the landfill cell (Table 1).
HTAW-3: Landfilled MSW, in the presence of moisture, undergoes microbial decomposition and this causes change in the geotechnical properties of MSW. To simulate this waste condition,
the influence of DOD and overburden stress on the geotechnical
properties of MSW were taken into account. To represent a typical state of degradation, a linear variation in the DOD and geotechnical properties of the MSW with depth was considered. It is
assumed that the topmost layer had the geotechnical properties
of fresh MSW (DOD=0 %), while the bottommost layer was
nearly completely decomposed (DOD=95%). This approach has
previously been adopted elsewhere (Reddy etal., 2011;
Sivakumar Babu etal., 2010). Based on the DOD values, the
geotechnical properties MSW at different depths were estimated
and are summarized in Table 1.
Unsaturated hydraulic parameters were kept constant for all
MSW cases and followed the values given by Breitmeyer and
Benson (2011) to evaluate the effect of two-phase flow (Table 2).
Unsaturated hydraulic properties of the MSW were not varied
with respect to the depth because: (1) very little published information is available on the evolution of unsaturated hydraulic
properties of MSW as a function of overburden pressure; and (2)
unsaturated hydraulic properties have a relatively small impact
on the key design parameters at steady-state conditions (Hayder
and Khire, 2005; Reddy etal., 2013).

190

Initial and boundary conditions and

model input parameters
Mechanical boundary conditions were applied by fixing the base
in both horizontal and vertical directions, so that the lateral and
vertical deformations of the landfill at the base were zero. The
lateral deformation was restrained on the right side boundary of
the model, whereas the side slope was free to move in both directions and the top boundary was free to move only in the vertical
direction. Hydraulic boundary conditions were taken into consideration by fixing the pore gas pressure and seepage at the top
boundary and at the side slope. The pore gas pressure was atmospheric at the seepage boundary, which was impermeable to the
liquid as long as the liquid pressure (water pressure) was negative: the gas pressure was taken as zero at boundary nodes where
the condition was not met (ITASCA Consulting Group, 2011).
The right-side boundary and the bottom of the landfill model
were considered to be impermeable (i.e. free pore pressures and
free saturation).
All grid points were initially free to vary based on the net
inflow and outflow from the neighbouring zones. Pore-water
pressure was fixed to zero at the leachate collection-and-removal
system to represent the drainage layer. The pore gas pressures
were fixed to be zero initially at all grid points in order to establish initial mechanical equilibrium. The initial pore-water pressure was calculated based on the initial gas pressure by default.
Thereafter, the gas pressures were set to vary for different flow
conditions (ITASCA Consulting Group, 2011). The initial waste
saturation of 40% at all grid points and an initial porosity of 40%
at all zones were considered.

Model simulations
Two-phase flow modelling, presented in this study, was validated
based on the published studies using single-phase flow modelling
and slope stability analysis under simplified conditions (e.g.
HAW). The effects of continuous elevated injection pressures on
the stability of bioreactor landfill slope were modelled and the
results were compared with the single-phase flow study in terms
of: (1) FOS vs. time, and (2) flow rate vs. time. For the validation
purpose, continuous injection pressures of 49, 98, and 147 kPa
were considered. In addition, the sensitivities of MSW geotechnical properties on the stability of landfill slope were analysed
and validated with a two-phase flow model; these results are presented elsewhere (Giri and Reddy, 2013).
All simulations were carried out using different continual
injection pressures to examine transient leachate distribution
until the steady state was reached or the injection time period for
which the landfill slope design became unacceptable (i.e.
FOS<1.5), whichever occurred first. Continuous leachate was
added with injection pressures of 49, 98, 147, and 196 kPa, and
these pressures were calculated based on water column heads of
5, 10, 15, and 20 m, respectively.
The results were obtained in terms of wetted area (moisture
distribution), pore-water and capillary pressures, and FOS with

Waste Management & Research 32(3)

injection times for different MSW conditions. In order to investigate the effect of HTAW, the results obtained for HTAW-2 (a realistic and representative of field conditions) were compared with
the results of HAW and HTAW-1. Since, for these three waste
conditions (HAW, HTAW-1, and HTAW-2), it was assumed that
the MSW is a freshly placed waste with zero DOD throughout the
landfill, the results of HTAW-2 were also separately compared
with HTAW-3 (most representative of field conditions), as this
condition takes into consideration of variation of geotechnical
properties of MSW due to DOD.

Results and discussion

Model validation
Figure 2a shows the effects of continuous injection pressures on
the slope stability. For validation purposes, the results obtained
from the present study are compared with the published results.
Xu etal. (2012) reported a reduction in FOS from 2.05 to 2.04 at
a continuous injection pressure of 49 kPa for the injection period
of 10 years, whereas the present study resulted in FOS reduction
from 2.05 to 2.02 for the same duration. As injection pressure
increased, leachate spread more in laterally than vertically downward due to the waste anisotropy. This resulted in larger waste
saturation near the trench location and to an increase in pore fluid
pressures. Consequently, FOS of the slope model reduced with
injection time. A continuous injection pressure of 98 kPa for 10
years led to reduced FOS from 2.05 to 1.98 (present study) and
approximately 2.05 to 2.00 (Xu etal., 2012), respectively.
Similarly, a continuous injection pressure of 147 kPa for 10 years
caused a decrease in FOS from 2.05 to 1.95 (Xu etal., 2012) and
2.05 to 1.94 (present study).
The determination of unit outflow rates with time due to different injection pressures was carried out in the present study
(Figure 2b) and compared with the results reported by Xu etal.
(2012). It was clear that the effects on unit flow rates due to continuous leachate injection at pressure range used were very limited and thus could be ignored.

Moisture distribution
Figure 3 shows the saturation contours for different MSW conditions with continuous leachate injection pressure of 196 kPa at
the end of 10 years of injection. For HAW, depending upon the
injection pressures, steady-state flow was attained with different
time periods of injection. The time needed to reach the steadystate condition was approximately 2140, 2320, 2410, and 2550
days for the injection pressures of 196, 147, 98, and 49 kPa,
respectively. It was assumed that the steady-state condition was
reached when the injected leachate was equal to the total leachate
seeping out through the landfill cell at the base and collected at
the leachate collection-and-removal system (i.e. when the saturation profile was perpendicular to the leachate collection-andremoval system); whereas no steady-state flow condition was
achieved for any of the aforementioned HTAW even after 10

Giri and Reddy

191

Figure 2. Model validation: effects of different injection pressures for homogeneous-anisotropic MSW for (a) factor of safety
vs. time and (b) outflow rate vs. time.

years of continuous leachate additions, since the saturated

hydraulic conductivities in successive deeper layers of the MSW
for all HTAW were significantly lower than that of HAW.
Furthermore, the leachate tended to migrate more laterally than
vertically downward and tried to seep out through side slope, primarily due to high saturated hydraulic conductivity in lateral
direction (kh=10 kv).
This study aimed at achieving a wetted area of MSW corresponding to 60% or higher degree of saturation, as recommended
by Alternative Landfill Technologies Team (2006). Figure 4
shows the evolution of wetted area of MSW with injection time
for different MSW conditions. The wetted area increased with
increase in injection pressure due to more migration of leachate
throughout the MSW. For continuous leachate injection with a
relatively high pressure of 196 kPa, the wetted area after 30 days
injection was 44 and 59% higher in HAW and HTAW-1, respectively, and 42% lower in HTAW-3 compared with HTAW-2. This
could be attributed to larger migration of leachate in successive

deeper layers of MSW in HTAW-1 and HAW than in HTAW-2

(Figure 3). However, for HTAW-3, the lateral spread of leachate
was limited and found to be lower than in HTAW-2, due to lower
saturated hydraulic conductivity of HTAW-3 than HTAW-2. This
led to a limited rate of leachate migration in the MSW and
resulted in a smaller wetted area. In addition, for shorter injection
duration, there would be large differences in the wetted areas for
different MSW conditions, and it is necessary to account for
HTAW-2 (a realistic field condition) for various leachate operations and designs in bioreactor landfill. A similar trend was
observed for other three injection pressures (49, 98, 147 kPa)
during a 4-week injection period for different MSW conditions.
Figure 5 summarizes the maximum wetted area for different
MSW conditions using different injection pressures after 10 years.
For any injection pressure, the maximum wetted area was in the
following order: HTAW-1>HAW>HTAW-2>HTAW-3. This signified the presence of higher wetted area (moisture distribution)
under HTAW-1 and HAW for different injection pressures and

192

Figure 3. Saturation contours (leachate distribution) after

10 years with continuous injection pressure of 196 kPa for (a)
HAW, (b) HTAW-1, (c) HTAW-2, and (d) HTAW-3.

indicated that, as the injection pressures were lowered, the wetted

areas in HTAW-1 and HAW increased noticeably. Hence, HTAW,
with varied unit weight and saturated hydraulic conductivity, must
be considered for a reliable and optimal design of bioreactor landfills incorporating HTs, as the simplified HAW accounted for the
larger distribution of leachate in landfills. In addition, HTAW-1
resulted in a higher wetted area with respect to HTAW-2, whereas
HTAW-3 lowered it because of the DOD varying with depth.

Pore fluid pressures

The generation and distribution of pore-water and capillary
pressures and the degree of saturation for different waste conditions are shown in Figures 6 and 7. The results are shown for the
first 30 days of injection. These observations were made at a
location 5 m left of the HT and at an elevation of 30 m from the

Waste Management & Research 32(3)

base. The results show that, due to continuous leachate recirculation over time, the degree of saturation increased from the
initial 40% until it became fully saturated. Capillary pressure,
which was 17 kPa maximum initially, implying higher initial
pore gas pressure than pore-water pressure, reduced gradually
with time as the leachate was continuously injected and became
zero once the MSW was fully saturated. However, the porewater pressure increased continually due to leachate injection.
For the high injection pressure of 196 kPa over 30 days, the
developed pore-water pressure at 5 m left of the trench was
approximately 8% lower in HAW than in HTAW-2, as the relatively lower hydraulic conductivity of the successive deeper layers of the MSW in HTAW-2 resulted in generation of higher
pore-water pressures than in simplified HAW. However, the
developed pore-water pressure in the case of HTAW-1 was
approximately 18% higher than HTAW-2, since the saturated
hydraulic conductivities of HTAW-1 in the deeper layers of the
MSW were lower than HAW, and HTAW-2 and, therefore,
resulted in a maximum pore-water pressure. Furthermore, the
developed pore-water pressure in the case of HTAW-3 was
approximately 43% lower with respect to HTAW-2.
A similar trend was observed for injection pressures of 147,
98, and 49 kPa. For example, using a low injection pressure of 49
kPa, during 4 weeks, the pore-water pressure was estimated to be
48% lower in HAW, while 37% higher in HTAW-1, when individually compared with respect to HTAW-2 (Figure 7).
Furthermore, pore-water pressure was 133% lower in HTAW-3
than in HTAW-2. Therefore, the results imply that it is critical to
consider the effects of real landfilled waste conditions (i.e.
HTAW-2), otherwise the developed pore-water pressures would
significantly be lower in simplified HAW. In addition, as the
injection pressure reduced (from 196 to 49 kPa), the relative
difference in pore-water pressure increased considerably across
different MSW cases, so the effect of heterogeneity and anisotropy of MSW must be assessed.
Similarly, the capillary pressures were also affected under different waste conditions. For example, with injection pressure of
49 kPa over 2 weeks, the developed capillary pressure in HAW
was 28.5% lower than in HTAW-2. In addition, the capillary
pressure in HTAW-3 was higher by approximately 21% than in
HTAW-2. Similar relationships were observed for other injection
pressures. However, as the injection pressure and duration
increased, the capillary pressure subsequently reduced due to the
continual increase in leachate saturation.
Figure 8 summarizes the maximum developed pore-water
pressure for different MSW conditions after 10 years of continual
injection. For all injection pressures, the maximum developed
pore-water pressure was in the following order: HTAW-1>HTAW2>HTAW-3>HAW. In the case of HAW, the pore-water pressures
were significantly reduced and found to be the lowest amongst all
MSW conditions due to the attainment of steady-state conditions
at different injection pressures. Therefore, simplified HAW
resulted in lower pore-water pressures than real field conditions,
and this scenario may lead to unreliable and unsafe designs since,

193

Giri and Reddy

Figure 4. Evolution of wetted area with time under elevated injection pressures for different MSW conditions: (a) HAW; (b)
HTAW-1; (c) HTAW-2; and (d) HTAW-3.

difference in maximum developed pore-water pressure increased

with decrease in injection pressures (from 196 to 49 kPa) for all
MSW conditions.

Slope stability analyses

Figure 5. Comparison of maximum wetted area for different

MSW conditions under different injection pressures after 10
years.

in real field conditions, pore-water pressures would be much

higher than in the case of simplified HAW. Furthermore, the
developed pore-water pressure was highest in HTAW-1 due to
relatively low permeability of deeper layers of the MSW. Also, the

The stability of the bioreactor landfill slope was evaluated in

terms of FOS with injection time for different MSW conditions.
Baseline (no leachate injection) FOS was computed to be 2.05,
2.06, 2.11, and 1.88 for HAW, HTAW-1, HTAW-2, and HTAW-3,
respectively. The variation in the baseline FOS is mainly due to
the varied geotechnical properties of MSW.
As shown in Figures 9 and 10, increase in injection pressure
during 10 years resulted in lowered FOS for all MSW conditions, primarily due to excessive pore fluid pressures. The
influence of pore fluid pressures was minimal in HAW due to
attainment of steady-state flow conditions and this led to relatively higher values of FOS during leachate injection. Hence, it
can be interpreted that bioreactor landfill designs incorporating
HAW would be unreliable and nonconservative, as the values
of FOS computed for different injection pressures were 415%
higher in HAW than in the case of real field conditions

194

Waste Management & Research 32(3)

Figure 6. Evolution of pore-water and capillary pressures

and saturation over 30 days with continuous injection
pressure of 196 kPa for (a) HAW; (b) HTAW-1; (c) HTAW-2; (d)
HTAW-3.

Figure 7. Evolution of pore-water and capillary pressures

and saturation over 30 days with continuous injection
pressure of 49 kPa for (a) HAW; (b) HTAW-1; (c) HTAW-2; (d)
HTAW-3.

(i.e. HTAW-2). Maximum pore fluid pressures were observed

in HTAW-1 due to relatively low hydraulic conductivities of
deeper layers of the MSW. These higher pore pressures consequently yielded significantly lowered FOS for HTAW-1, especially at a high injection pressure of 196 kPa, for which the
design of bioreactor landfill slope became unacceptable (i.e.
FOS=1.481.5) after 10 years of continuous leachate injection
(Figure 9a). Even though for different injection pressures (49
196 kPa), the values of FOS were 210% lower in HAW than
HTAW-2 due to higher developed pore pressures and resulted
in unaccepted landfill designs (i.e. FOS<1.5), it is recommended to consider HTAW-2 rather than HTAW-1 for designs
and various leachate operations in bioreactor landfill, as the
MSW unit weight in HTAW-1 remained constant throughout
the depth and, hence, did not represent the most realistic field
conditions.
The MSW in HTAW-3 yielded FOS values in the range of
311% higher than in HTAW-2 for different injection pressures,

even though the pore fluid pressures were relatively lower in

HTAW-3 than in HTAW-2. This was most likely due to varying
DOD and shear strength of the MSW with depth for HTAW-3,
wherein the reduced friction angles (from 35 to 25.5; Table 1)
seemed to have more influence on the FOS computations than the
increased cohesion values (from 15 to 29 kPa; Table 1) and led to
lower FOS values than HTAW-2 during leachate injection.
However, further research is required to establish adequate relationship between degree of decomposition and geotechnical
properties of landfilled waste in order to comprehensively examine the effects of HTAW-3.

Conclusion
In this study, a numerical two-phase flow modelling was used to
predict the moisture distribution, pore-water and capillary pressures, and their impacts on stability of bioreactor landfill slopes.
Numerous simulations were performed to investigate the effect of

Giri and Reddy

heterogeneous-anisotropic nature of MSW during leachate
injection.
When estimating slope stability under leachate injection:

Figure 8. Maximum developed pore-water pressure for

different MSW conditions under different injection pressures
after 10 years of continuous leachate injection.

195
Steady-state flow was possible only in the case of homogeneous and anisotropic MSW (HAW), wherein the time to
reach steady-state flow varied with injection pressures,
with higher injection pressure resulting in a relatively
faster attainment of the steady state. The steady-state flow
in HAW led to relatively lower pore fluid pressures and
higher FOS during leachate recirculation. Thus, simplified
HAW represented the nonconservative and unreliable
design approach and leachate operations in bioreactor
landfill.
Steady-state flow was not achieved in the case of heterogeneous and anisotropic waste (HTAW) for different injection
pressures during 10 years of continuous injection, because, in
the case of HTAW, the spread of leachate was more in the
lateral direction than in the vertical downward direction, and
hence, the leachate tried to seep through the side slope of
landfills.
Neglecting waste heterogeneity in depth (i.e. simplified
homogeneous MSW) resulted in overestimation of the MSW
wetted area (moisture distribution) by 540%, underestimation of the developed pore-water pressure by 848% for

Figure 9. Evolution of factor of safety with injection time under different MSW conditions for different injection pressures: (a)
196 kPa; (b) 147 kPa; (c) 98 kPa; (d) 49 kPa.

196

Waste Management & Research 32(3)

Funding
This project was supported by the US National Science Foundation
(grant CMMI 0600441).

References

Figure 10. Factor of safety for different MSW conditions

under different injection pressures after 10 years (baseline:
no leachate injection).

shorter duration (i.e. 30 days) and by 2257% for longer

duration (i.e. 10 years) for different injection pressures (49
196 kPa), and consequently overestimation of the values of
FOS for bioreactor landfill slope by 415% over 10 years. In
addition, the capillary pressures for different MSW conditions were highest at the beginning of the leachate injection
(1520 kPa) and decreased to zero over time due to continuous leachate injection.
Neglecting the variations in MSW saturated hydraulic conductivity and unit weight resulted in overestimation of the
MSW wetted area and developed pore-water pressures by
659% and 516%, respectively, for different range(s) of
commonly employed injection pressures (49196 kPa) over
10 years. Subsequently, the values of FOS were underestimated by 210%.
Neglecting the different levels of waste degradation
resulted in underestimation of the MSW wetted area and
pore-water pressures by 2742% and 43133%, respectively, for different range of continuous injection pressures
over 10 years. The stability of landfill slope was overestimated by 310%.
Overall, it is concluded that heterogeneous and anisotropic
nature of MSW greatly influences the distribution of leachate, the
generation and distribution of pore fluid pressures in landfilled
waste, and the stability of bioreactor landfill slope during continuous leachate injection. Additionally, it is recommended that
heterogeneous and anisotropic MSW with varying unit weight
and saturated hydraulic conductivities with depth must be considered for designs and leachate operations as it provides a safer,
critical, and more reliable approach than simplified homogeneous and anisotropic MSW.

Declaration of conflicting interest

The authors declare that there is no conflict of interest.

Alternative Landfill Technologies Team, Interstate Technology and

Regulatory Council. (2006). Characterization, design, construction, and
monitoring of bioreactor landfills. Washington DC: ITRC.
Barlaz M, Ham R and Schaefer D. (1992) Microbial, chemical and methane production characteristics of anaerobically decomposed refuse with
and without leachate recycling. Waste Management and Research, 10,
257267.
Breitmeyer RJ and Benson CH. (2011) Measurement of unsaturated hydraulic properties of municipal solid waste. In: Han J and Alazamora D. (eds),
GeoFrontiers 2011: advances in geotechnical engineering, GSP no. 211.
Reston, VA: ASCE. pp. 14331442.
Chugh S, Clarke W, Pullammanappallil P, etal. (1998) Effect of recirculated
leachate volume on MSW degradation. Waste Management and Research
16: 564573.
Dawson EM, Roth WH and Drescher A. (1999) Slope stability analysis by
strength reduction. Gotechnique 49: 835840.
Eid HT, Stark TD, Evans WD, etal. (2000) Municipal solid waste slope failure I: waste and foundation soil properties. Journal of Geotechnical and
Geoenvironmental Engineering 126: 397407.
Gharabaghi B, Singh MK, Inkratas C, etal. (2008) Comparison of slope
stability in two Brazilian municipal landfills. Waste Management 28:
15091517.
Giri RK and Reddy KR. (2013) Two-phase flow modeling of leachate injection
effects on stability of bioreactor landfill slopes. Proceedings of the Air and
Waste Management Association, 106th Annual Conference and Exhibition.
Pittsburgh, PA: Air and Waste Management Association. pp 117.
Hayder MM and Khire MV. (2005) Leachate recirculation using horizontal trenches in bioreactor landfills. Journal of Geotechnical and
Geoenvironmental Engineering 131: 837847.
ITASCA Consulting Group. (2011) FLAC Fast Lagrangian Analysis
of Continua, ITASCA Consulting Group manuals. Minneapolis, MN:
ITASCA.
Kavazanjian Jr E and Merry SM. (2005) The 10 July 2000 Payatas landfill
failure. In: Proceedings of Sardinia-10th International Symposium Waste
Management and Landfill, paper no. 431 (CD ROM). Cagliari, Italy:
International Waste Working Group (IWWG), CISA Publisher.
Koerner RM and Soong TY. (2000) Leachate in landfills: the stability issues.
Geotextiles and Geomembranes 18: 293309.
Mitchell JK, Seed RB and Seed HB. (1990) Kettleman Hills waste landfill slope failure. I: liner-system properties. Journal of Geotechnical
Engineering 116: 647668.
Reddy KR, Hettiarachchi H, Gangathulasi J, etal. (2011) Geotechnical properties of municipal solid waste at different phases of degradation. Waste
Management 31: 22752286.
Reddy KR, Kulkarni HS and Khire MV. (2013) Two-phase modeling of leachate recirculation using vertical wells in bioreactor landfills. Journal of
Hazardous, Toxic, and Radioactive Waste 17: 272284.
Reddy KR, Hettiarachchi H, Parakalla N, etal. (2009a) Geotechnical
properties of landfilled municipal solid waste under short-term leachate recirculation operations. Waste Management and Research 27:
578587.
Reddy KR, Hettiarachchi H, Parakalla N, etal. (2009b) Hydraulic conductivity of MSW in landfills. Journal of Environmental Engineering
135: 17.
Reinhart DR, McCreanor PT and Townsend T. (2002) The bioreactor
landfill: its status and future. Waste Management and Research 20:
172186.
Sivakumar Babu GL, Reddy KR, Srivastava A, etal. (2010) Reliability
analysis of municipal solid waste landfill slopes. In: Proceedings
from Sixth International Congress on Environmental Geotechnics,
812 November 2010. New Delhi, India: Tata McGraw Hill Education
Private Limited, pp.17111716.

Giri and Reddy

Tchobanoglous G, Theisen H and Vigil SA. (1993) Integrated solid waste
management, engineering principles and management issues. NewYork:
McGraw-Hill.
van Genuchten MT. (1980) A closed-form equation for predicting the hydraulic conductivity of unsaturated soils. Soil Science Society of America
Journal 44: 892898.
Xu Q, Tolaymat T and Townsend TG. (2012) Impact of pressurized liquids addition on landfill slope stability. Journal of Geotechnical and
Geoenvironmental Engineering 138: 472480.

197
Yeiller N, Hanson JL and Liu WL. (2005) Heat generation in municipal
solid waste landfills. Journal of Geotechnical and Geoenvironmental
Engineering 131: 13301344.
Zekkos D, Bray J, Kavazanjian E, etal. (2006) Unit weight of municipal
solid waste. Journal of Geotechnical and Geoenvironmental Engineering
132: 12501261.
Zhan T L, Chen Y M and Ling WA. (2008) Shear strength characterization of
municipal solid waste at the Suzhou landfill, China. Engineering Geology
97: 97111.