Alexandria Engineering Journal (2019) xxx, xxx–xxx

H O S T E D BY
Alexandria University

Alexandria Engineering Journal

www.elsevier.com/locate/aej
www.sciencedirect.com

ORIGINAL ARTICLE

Assessment of slope instability of canal with

standard incomat concrete-filled geotextile
mattresses lining
Mohamed Galal Eltarabily a,b,*, Hossam Eldin Moghazy c, Abdelazim M. Negm d

a
Civil Engineering Department, Faculty of Engineering, Port Said University, 42523 Port Said, Egypt
b
Kearney Agricultural Research and Extension Center, University of California, Parlier, CA 93648, USA
c
Irrigation Engineering and Hydraulics Dept., Faculty of Engineering, Alexandria University, 21527 Alexandria, Egypt
d
Water and Water Structures Engineering Department, Faculty of Engineering, Zagazig University, Zagazig 44519, Egypt

Received 2 October 2019; revised 30 October 2019; accepted 16 November 2019

KEYWORDS Abstract Geosynthetics polymeric materials are recently widely used for geotechnical problems of
Ismailia canal; canals specifically seepage losses with cost-effectiveness, low environmental impact, and quantifi-
Slope stability; able performance. While the geotextile is one of the geosynthetics, incomat filled concrete mattress
Geotextile; is the standard and most famous polyethylene (PE) and polyamide (PA) double woven type of geo-
Incomat concrete mattress; textiles. Its main functions are as a liner and/or controlling erosion of soil that creates changes of
SLOPE/W permeability which could affect the stability of the side slopes of canals. Among irrigation canals in
Egypt, Ismailia canal (129.5 km length) is considered the worst case of seepage problems and ero-
sions. The Ministry of Water Resources and Irrigation (MWRI) implemented preliminary technical
studies to locate the most critical sections along the canal which seepage losses are the maximum.
Lining samples (100 m2, area) by standard incomat were executed in 2016 where four sections of
lining with a total length of 61 km were defined. This research assesses the effect of applying the
geotextile liner on the slope instability of Ismailia canal at the different proposed four locations
of lining where the canal geometry and soil properties are variables. SLOPE/W module of GeoStu-
dio finite element software was used to model the simulation after a clear definition of the soil and
geotextile physical properties. The factor of safety, Fs of slope stability for the normal condition of
the canal (without lining) ranges from 1.68 to 2.23. The lowest Fs (1.421) was obtained for the first
part (km 17.5–28.0) of the second lining section of the canal (km 17.5–38.0) where bed width equals
49 m, maximum and minimum water depth equals 5.79 m and 5.05 m where the soil type is fine sand
and the side slopes are 2:1. After implementing the lining, the minimum factor of safety was 1.421
for the same previous location with a 15.57% reduction. The maximum reduction in Fs was
obtained for the second part (km 75.0–92.5) of the fourth lining section (km 72.0–92.5) while it
records 21.61% with a factor of safety 1.491. Those two parts are critically affected after imple-
menting the lining. However, the maximum reduction in Fs was not exceeded 5% for all sections
with sandy clay soil where the Fs did not fall below 1.70. While the geotextile is significant in seepage

* Corresponding author. Tel.: +1 (559)672-0631; +201222592623.

E-mail addresses: [email protected], [email protected] (M.G. Eltarabily), [email protected] (H. Eldin
Moghazy), [email protected] (A.M. Negm).
Peer review under responsibility of Faculty of Engineering, Alexandria University.
https://doi.org/10.1016/j.aej.2019.11.010
1110-0168 Ó 2019 Faculty of Engineering, Alexandria University. Production and hosting by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Please cite this article in press as: M.G. Eltarabily et al., Assessment of slope instability of canal with standard incomat concrete-filled geotextile mattresses lining,
Alexandria Eng. J. (2019), https://doi.org/10.1016/j.aej.2019.11.010
2 M.G. Eltarabily et al.

controlling, particularly at locations where the soil type is coarse sand, specifications and require-
ments of the thickness uniformity of the mattresses should be achieved for the concrete filling pro-
cess during the construction and measurements of long-term durability could be considered to
sustainably function.
Ó 2019 Faculty of Engineering, Alexandria University. Production and hosting by Elsevier B.V. This is an
open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction The problem of slope instability of canals is widely simu-

lated by GeoStudio, geotechnical software based on the limit
The most frequently encountered problems of irrigation canals equilibrium principles which incorporate a finite element pack-
networks are the seepage that causes erosion and instability of age developed by [11]. To understand the stability of side
side slopes [1]. While the seepage losses mainly depend on the slopes with a combination of the seepage discharge from the
stream geometry and hydraulic conductivity of the soil, the canal, coupled modules of SEEP/W–SLOPE/W have been
highest values obtained for the coarse soil, sand and gravel increasingly used [12–14]. SLOPE/W uses various methods to
than the fine soil [2]. When seepage occurred and the flow compute the factor of safety such as the ordinary slice method,
velocity exceeded the maximum limiting velocity, the hydraulic Bishop’s method, Janbu’s method, Spencer’s method, Sarma’s
gradient transcended the critical gradient and soil particles method (vertical slice only), Lowe Karafiath, and Morgenstern
start to move outward the slope with the water flow thus leads Price method. Both deterministic and probabilistic soil param-
to slope instability until failure occurred [3]. However consid- eters can be assigned to perform the slope stability analysis.
ering these technical problems in the design of canals, some However, SLOPE/W lacks the dynamic hydrological modeling
deficiencies could be obtained during operations and other of pore water pressure therefore; simulated seepage from
alternatives of solutions could be included [4]. Geosynthetics SEEP/W is directly imported into SLOPE/W simulation.
polymeric materials are one of those solutions which their Among the various above-mentioned methods of limit equilib-
main functions are as liners and/or controlling erosion and rium, the Morgenstern Price method [15] is widely used to
slope reinforcements whilst their execution is in dry conditions compute the factor of safety. It satisfies both force and
during the construction of the canal or underwater while the moment equilibrium, applicable to the general shape of slip
canal is in operations. Geosynthetics can be categorized based surfaces, and allows the use of special variable inter-slice force
on their geotechnical functions to geomembranes, geotextiles, function so that its inclination can vary along slip surface [14].
geogrids, genets, geocells, geomats, and the combination of For achieving a comprehensive assessment of the geo-
these products into geocomposites [5]. synthetic textile technique after locating the most critical four
Among the irrigation system of canals and drainage net- lining sections along the Canal by MWRI, the influence of this
works in Egypt, Ismailia canal is experiencing severe problems lining method on the slope instability of the canal at the differ-
of poorly functioning, seepage, erosions, and slopes instability ent locations could be investigated and compiled with the final
[6]. The development of the canal has been continuously imple- impact assessment report of the project. Consequently, the
mented by the Ministry of Water Resources and Irrigation [7] objective of this research is to identify the effect of using the
for increasing the canal capacity and performance for sustain- standard incomat concrete-filled geotextile mattresses as the
able management with the expansion of the irrigated area by lining technique of seepage controlling of Ismailia canal on
the canal [8]. While the implementation of the development slope instability at the four proposed lining locations by
plan of the canal through three phases of increasing the dis- MWRI including the hydrogeological conditions and the vari-
charge by widening the cross-sections along the canal length, ation of the canal geometry along its length.
negative impacts of increasing the seepage losses were occurred
because of removing the silty layers without subsequent 2. Description of the project sites
replacement [9]. Later, MWRI with the financial support of
the African Development Bank (AFDB) initiated a project in Ismailia canal is a part of the irrigation canals network in the
2016 for canal rehabilitation after the comprehensive feasibil- eastern Nile Delta while it serves 320,000 ha of area for agri-
ity studies that were performed in 2014 [10]. While four sec- culture. It is considered as a main branch canal where its
tions with 61 km total length along the canal were defined as intake on the Nile River at Cairo until reaching Ismailia city
critical for seepage and erosions, the first component of the to the northeast and outlets on the Suez Canal with a total
three objectives of the project is for dredging and lining the length of 129.5 km, while the DEM map of the whole eastern
stretch between km 72 and km 92.5. These four defined lining Nile Delta area is shown in Fig. 1-a. The previous environmen-
sections are as following; from km 7.0 to km 13.0, from km tal and social impact assessment (ESIA) studies showed the
17.5 to km 38.0, from km 44.0 to km 58.0, and from km most deteriorated stretches along the canal that were catego-
72.0 to km 92.5 while the km numbering starts from the inlet rized as the first priority of the project components (rehabilita-
of the canal at Cairo and ends by 129.5 km at Ismailia city. tion stage) [10]. The four predefined locations for dredging and
Three prevention techniques including Geo-Membrane, Geo- lining of the canal are shown in Fig. 1-b. While Ismailia canal is
Synthetic Textile, and Geo-Synthetic Clay were selected and located on the southeastern edge of the Nile Delta, different geo-
executed on a pilot basis of 100 m2 area for each and the most logical formations were merged. At about km 70, the canal lies in
effective option will be selected from the environmental and the old lands of the fluviatile Nile Delta deposits and follows a
socio-economic aspect. depression to Ismailia governorate [16]. After the Km 115 and

Please cite this article in press as: M.G. Eltarabily et al., Assessment of slope instability of canal with standard incomat concrete-filled geotextile mattresses lining,
Alexandria Eng. J. (2019), https://doi.org/10.1016/j.aej.2019.11.010
Assessment of slope instability of canal 3

Fig. 1 (a) Digital elevation model for the whole Nile Delta showing Ismailia canal and other irrigation canals and drains network. (b)
Ismailia canal from its inlet at Cairo until it reaches Ismailia governorate showing the four lining sections and their length.

until Ismailia, the canal lies in Sandy desert formation with ter- sections were generated (two for each lining part) where they
tiary Pliocene-Miocene sands and gravel [17]. The pre- namely from 1 to 4 with A, and B letters and the surrounding
feasibility studies show that in the current situation without soil types are shown (Fig. 3a–h). The length of each segment
implementing the project, the canal does not cover the water (part) along the four lining sections is shown in Table 1.
needs for agriculture as well as for drinking water supplies Four lining alternatives were compared during the pre-
and industries [10]. Two main hydrological zones of Ismailia feasibility studies of using clay mattresses in sections of the
canal present the study area as following; zone A of confined canal with complete dewatering, laying plates underwater, lay-
aquifer impermeable cap layer covers the aquifer where the soil ing geomembrane under wet condition, and the final alterna-
types are clay and/or sandy clay [18]. Zone B of unconfined tive of pumping clay/cement into a frame, underwater. Three
condition where the aquifer is phreatic and the soil types that lining technologies were executed on a pilot-scale of using
mainly represent the aquifer are coarse sand, and sand and geosynthetic textile in December 2014, geomembrane tech-
gravel (Fig. 2). This top layer as a part of the whole conceptual nique in February 2016, and geo-synthetic clay liner in March
model coverage the study area was generated from forty-five 2016. In this research, the chosen studied technique was the
boreholes data by the MODFLOW model and the lithostratig- geotextile because of its feasibility and shortest duration of
raphy of the aquifer was obtained as well [19]. While the canal construction (3–4 years) with the lowest cost with comparison
geometry and discharge changes along its length and the to other alternatives [10]. Fig. 4a–c shows the construction
hydrogeological conditions are different [20], eight cross- process of HUESKER geo-synthetic textile that is assigned
in the model simulation as a liner of seepage control and
surcharge load over the canal slopes. The polyethylene (PE)
and polyamide (PA) double-woven with geotextile formwork
incomat mattress ensure a constant thickness concrete
cross-section while its adaptability gives it a clear edge over
all conventional concrete solutions. Fig. 5 summarizes the
bed levels, high and low water levels, maximum and minimum
discharge along the 129.5 km canal length.

3. Numerical modeling

In order to model the real lining conditions underwater during

the canal operation, the SEEP/W module was first initiated to
calculate the seepage losses and water flow for the eight differ-
ent cross-sections along with the four lining parts. The tran-
sient state SEEP/W model was performed for a one-year
Fig. 2 The top layer of the aquifer studied area that served by simulation to include the variation of the canal discharge (min-
Ismailia canal showing the clay and sandy clay parts (confined imum and maximum during the winter and summer months).
conditions) and the fine sand and coarse sand (unconfined). One month time step was defined and the maximum water

Please cite this article in press as: M.G. Eltarabily et al., Assessment of slope instability of canal with standard incomat concrete-filled geotextile mattresses lining,
Alexandria Eng. J. (2019), https://doi.org/10.1016/j.aej.2019.11.010
4 M.G. Eltarabily et al.

Fig. 3 The eight cross-sections of the canals while two sections for each lining part (a) for the 1st part of the 1st lining section (b) the 2nd
part of the 1st lining section (c) the 1st part of the 2nd lining section (d) the 2nd part of the 2nd lining section (e) the 1st part of the 3rd
lining section (f) the 2nd part of the 3rd lining section (g) the 1st part of the 4th lining section (h) the 2nd part of the 4th lining section.

Table 1 The four lining sections locations and the eight parts with detailed lengths along the canal.
Section Part (segment) From (km) –To (km) Part length (km) Section length (km)
1st 1st (A) 7.0–10.5 3.50 6.0
2nd (B) 10.5–13.0 2.50
2nd 1st (A) 17.5–28.0 10.50 20.5
2nd (B) 28.0–38.0 10.0
3rd 1st (A) 44.0–49.0 5.0 14.0
2nd (B) 49.0–58.0 9.0
4th 1st (A) 72.0–75.0 3.0 20.5
2nd (B) 75.0–92.5 17.5

Fig. 4 (a) After finishing the concrete pouring for the bottom layer and spreading the slopped mattress (b) elements components of the
standard incomat mattress used in the project (c) continuing filling the slopped part with concrete until the canal embankment.

Please cite this article in press as: M.G. Eltarabily et al., Assessment of slope instability of canal with standard incomat concrete-filled geotextile mattresses lining,
Alexandria Eng. J. (2019), https://doi.org/10.1016/j.aej.2019.11.010
Assessment of slope instability of canal 5

Fig. 5 Bed level, H.W.L, and L.W.L, maximum and minimum discharges along the canal after [20].

levels were assigned from May to October and the minimum simulation are shown in Table 2. After the 365 days transient
water levels were assigned for other months during the year. simulation of SEEP/W, the water table was obtained and
The shear strength and soil hydraulic parameters used in the included in the SEEP/W simulation. For the seepage modeling,
the soil-water characteristic curve (SWCC) function was
assigned from the GeoStudio Artificial Neural Network
Table 2 Shear strength and hydraulic parameters used for (ANN) database calculations based on the grain size distribu-
SLOP/W and SEEP/W models [22]. tion (Fig. 6-a). The soil hydraulic conductivity curve was esti-
Parameter Soil Type mated from the SWCC using Van Genuchten function [21]
(Fig. 6-b) and the residual moisture content as given by [22].
Sandy Fine Coarse
In order to limit the excessive-high negative pore water pres-
Clay Sand Sand
sure, 10 KPa suction was imposed as an initial condition to
Unit weight, c (kN m
3) 20.2 19.1 17.8 integrate the model with the real field condition of saturation.
0
Effective cohesion, c (Kpa) 20 2 1
0
The bottom boundary of the model domain was assigned as a
Angle of shearing resistance, u (°) 15 31 34
free drainage flux to express the seepage out of the domain to
Saturated moisture content, hs (cm3 0.425 0.250 0.190 the aquifer storage. The porewater pressure distribution is sim-
cm
3)
ulated by solving Darcy’s equation for unsaturated and satu-
Residual moisture content, hr (cm3 cm
3) 0.240 0.100 0.050
Hal. Hydraulic conductivity, Kx (m 0.85 1.10 78.00 rated flow conditions in each time step one month) for the
day
1) two-dimensional seepage calculations (Eq. (1)). Then,
Hal anisotropy, Khal/Kval (__) 0.10 0.20 0.20 SLOPE/W was paired to include the calculation of the water
table in the soil profile adjacent to the slop.

Fig. 6 (a) Soil water suction functions for the three soil types including in the seepage simulation (sandy clay, fine sand, coarse sand) (b)
the soil permeability curve for the three soil types from [21].

Please cite this article in press as: M.G. Eltarabily et al., Assessment of slope instability of canal with standard incomat concrete-filled geotextile mattresses lining,
Alexandria Eng. J. (2019), https://doi.org/10.1016/j.aej.2019.11.010
6 M.G. Eltarabily et al.
   
@ @H @ @H @H must lie on the perpendicular bisector of the straight line from
kx þ ky þ q ¼ m2w cw ð1Þ
@x @x @y @y @t the point of Xc to the slope toe, O where Xcc is the abscissa of
the critical slip surface circle center [26]. So, when Xcc is
where kx is the coefficient of permeability in the x-direction, ky
defined, then the critical slip center can be determined and
is the coefficient of permeability in the y-direction, H is the
both together (Xc, Xcc) gives the circular (failure) slope surface.
hydraulic head or total head, q is the applied flux at the
The ranges of the entry and exit search region were assigned
boundary, mw is the slope of the soil-water characteristic
5.0 m with 0.5 m interval; L and S interval were also assigned
curve, and cw is the unit weight of water.
0.5 m in the simulation. After the random generation of Xc,
In SLOPE/W simulation, the inter-slice force function is
and Xcc, and the evaluation of fitness for each iteration, the
varying only in the Morgenstern Price method so, it is consid-
algorithm terminates when satisfactory fitness level has been
ered more accurate than other methods thus, it was accounted
achieved which means the minimum factor of safety was
for modeling the slope stability in this study. The half-sine
reached [27]. In order to assess the effect of the lining project
inter-slice force function was used in the simulation since it
on the slope instability of the canal, the factors of safety were
tends to concentrate the inter-slice shear force towards the
compared for the eight simulations (2 segments for each lining
middle of the sliding mass while diminishes the inter-slice shear
section) for the current canal operation (with no lining) and
in the crest and toe areas. The modified Mohr-Coulomb soil
after implementing the lining project. The geotextile filled-
strength model [23] which includes the variation of shear
concrete mattress functionally acts as an additional load over
strength due to matric suction in unsaturated soil is expressed
the ground surface of the slope. 40 cm height of the surcharge
in Eq. (2). The simulation domain and geometry of the model
load was assigned as a uniform thickness of the mattress with
boundary are the same for the SEEP/W and SLOPE/W for
22 KN m
3.
the same specified cross-section with the same mesh size
(0.5 m  0.5 m).
4. Results and discussion
0 0
s ¼ c þ ðrn
ua Þtanu þ ðua
uw Þtanub ð2Þ
0
where s is the shear strength of unsaturated soil; c is the effec- The SEEP/W analysis was firstly run and the water table was
tive cohesion; ðrn
ua Þ is the net normal stress; rn is the total computed during the one-year transient state simulation while
normal stress; (ua
uw Þ is the matric suction;uw is the porewa- the initial water table was horizontal and the water levels in the
0 canal varies from the H.W.L. to the L.W.L for the summer
ter pressure; u is the angle of shearing resistance; and ub is the
and winter seasons. After solving the SEEP/W model,
angle expressing the rate of increase in shear strength relative
SLOPE/W was executed and the mode of failure and factors
to the matric suction when the soil approaches saturation,
0 of safties was recognized for the different sections along the
ub reaches an ultimate value of u [24].
canal for the case of no lining and after implementing the
While the reliability of the slope stability simulation results
geotextile.
mainly depends on the method of determination of the critical
slip surface, SLOPE/W has many techniques of defining the
4.1. Mode of failures of the current situation of the canal without
area of potential failure. The entry and exit method is chosen
lining
in the simulation where it depends on the genetic algorism, GA
which is a robust search technique that overcomes drawbacks
of other methods and results are satisfactory [25]. Whist, the While the canal passes through the topsoil layer, which its
GA steps of the solution includes mutation, selection, and depth covers the whole vertical extent of the simulated domain,
crossover, the two-dimensional circular slope failure is deter- along its path, the homogeneous soil profile was assigned for
mined by two variables Xc and Xcc (in Fig. 7) where Xc is each simulation region, (20 m, depth  50 m length  1 m
the abscissa of a point on slope top surface. When Xc is wide). Thus, the rotational mode with circular shape occurred
located, the critical slip surface circle center with abscissa Xcc for all eight segments and the toe failure was obtained where
the failure plane passes through the base of slope although,
the 5.0 m entry region was assigned from the toe to the back-
ward. Fig. 8a through 8 h show the failures of the eight seg-
ments for the four lining sections along with the Fs and the
radius of the circular slip. Although, the sections of the fine
sand and coarse sand have lower steep slopes than the sandy
clay’s sections (2:1 and 3:2, respectively), it is evident that
the Fs for sandy clay is higher than for the fine and coarse sand
even for the same relative geometry (vertical projection of the
slope = difference between top ground level and the bed level).
Firstly, the obtained factors of safety for the sandy clay
soils which vary between 1.760 and 2.316 are higher than the
factors of safety for the fine sand and coarse sand that varies
from 1.683 to 2.081. Additionally, the maximum factor of
safety, Fs was localized for the 5 km segment between km 44
and km 49 (1st part of the 3rd section) while the minimum
Fs occurred along the 10.5 km length from km 17.5 to km 28
Fig. 7 Description of the entry and exit method based on GA (the 1st part of the 2nd section). In addition to the mode of
technique for defining the failure surface. toe failure, it is clearly shown the significant difference in the

Please cite this article in press as: M.G. Eltarabily et al., Assessment of slope instability of canal with standard incomat concrete-filled geotextile mattresses lining,
Alexandria Eng. J. (2019), https://doi.org/10.1016/j.aej.2019.11.010
Assessment of slope instability of canal 7

Fig. 8 The critical slip surfaces showing the factor of safety and radius of the circular failure surface (a) (Sec1-A) (b) (Sec1-B) (c) (Sec2-
A) (d) (Sec2-B) (e) (Sec3-A) (f) (Sec3-B) (g) (Sec4-A) (h) (Sec4-B) for the current situation without lining.

Please cite this article in press as: M.G. Eltarabily et al., Assessment of slope instability of canal with standard incomat concrete-filled geotextile mattresses lining,
Alexandria Eng. J. (2019), https://doi.org/10.1016/j.aej.2019.11.010
8 M.G. Eltarabily et al.

slip volume and the radius of the circular plan of failure among slip surface reaches the top ground surface, allowing for the
the relatively fine soil (sandy clay) and the coarse soil (fine development of large displacements. While the computed shear
sand, and the coarse sand) Table 3. It clearly appears the sig- strength for the sandy clay soil based on the assigned shear
nificant variance of volume of the slip surfaces among the sand parameters is higher than for the sandy soil, more resistance
soil (large volume) and the fine and coarse sand soils which is of the soil and time until reach the total failure has occurred
low that reflects the narrow width of the slip near to the slope. thus the bigger shape of the slip surface with more extended
For the first four segments (Fig. 8 a-d) where the vertical pro- periphery are obtained.
jection of the slope is relatively the same (ranges around 8 m to Moreover, to understand the variation in the factor of
9 m) the volumes of the critical slip for the sandy clay soil are safety, Fs for the same soil type with the same degree of
much higher than for the coarse and fine sand. Additionally, slope, where the four sandy clay sections have slope 3:2
the radius of the slip surface for the coarse and fine sand are and the three sections of the fine sand with the 2:1 slope,
31.8 m and 28.0 m while they are 14.6 m and 15.4 for the sandy the following calculations were performed (Table 4). From
clay soil. the bed levels, the top ground levels, and the L.W.L
Thus confirms the variation of the mechanism of the failure (because the end of the year in December was assigned
propagation for the sandy clay soil than the fine and coarse the low water levels), the total vertical depth of the slope
sandy soils. The evolution of the slope failure is interpreted was calculated and the water depth as well. It is obviously
by the shear strain that initiates the creation of the local slip saying that the consequence of the factor of safety for the
surface in intact soil and increased until the general failure same soil type is directly related to the percentage of the
occurred [28]. More interpretation of the mechanism starts depth of water by the total vertical depth. Thus, the more
with local failure which firstly occurs in a limited zone at the percentage of water depth relative to the total depth, the
toe of the slope, where the peak strength is mobilized along more factor of safety obtained which means more added
with directions depending on the direction of local principal stability by the weight of the water volume acts perpendic-
stresses [29]. This process could generate small shear fissures ular to the slope [32].
(minor shears) in the direction of failure planes which are
not necessarily consistent with the formation of a continuous 4.2. Effect of geotextile filled-concrete mattress on the factor of
slip surface, therefore; at this stage, only finite strains are safety
allowed [30]. A slip surface can be developed only in a subse-
quent phase, as a consequence of further shear stress increase, After the installation of the geotextile mattress, the low perme-
causing a rotation of failure planes and formation of new able light concrete was filled with a uniform 40 cm thickness.
shears that merge into a unique principal shear [31]. Further Therefore, the assigned surcharge load of the mattress was
stress change can propagate the shear zone and develop the applied for the SLOPE/W simulation by considering the sub-
slip surface. In this stage, the global safety factor of the slope merged unit weight of the part of the mattress under the water.
remains larger than one, even if locally, it is equal to one [27]. The factor of safety and the moment equilibrium including the
The general slope failure (full failure) will occur only when the slip details; volume, length, radius for the eight segments were

Table 3 Factors of safety for the eight parts for the current situation of the canal and after implementing the lining, showing the
volume of the critical slip and the equilibrium moments, radius, and the longitudinal length from the start to the end of the circular
failure surface.
Section The current situation of the canal without lining
Fs Volume, V (m3) Length, l (m) Radius, R (m) Mstab./Mover. (KN m)
1st (6.0 km) (A) 2.088 53.549 18.288 15.417 9696.1/4641.8
(B) 2.227 47.001 17.278 14.664 8452.5/3794.9
2nd (20.5 km) (A) 1.683 31.461 21.298 31.793 9976.8/5928.6
(B) 2.081 25.173 18.753 28.073 8092.4/3889.3
3rd (14.0 km) (A) 2.316 42.192 16.187 13.613 7330.8/3165.2
(B) 1.760 62.889 20.109 17.195 12454.0/7076.2
4th (20.5 km) (A) 1.884 15.507 14.273 21.310 3344.6/1775.5
(B) 1.902 11.528 12.140 17.932 2092.0/1099.9
Section After lining with standard incomat geotextile
Fs Volume, V (m3) Length, l (m) Radius, R (m) Mstab./Mover. (KN m)
1st (6.0 km) (A) 2.015 41.597 16.822 14.757 8259.2/4099.4
(B) 2.117 35.693 15.851 14.019 7191.6/3397.7
2nd (20.5 km) (A) 1.421 27.751 20.757 31.458 10,064/7081.3
(B) 1.778 18.670 17.798 27.411 8162.1/4590.6
3rd (14.0 km) (A) 2.221 31.670 14.741 12.953 6116.8/2753.5
(B) 1.702 49.593 18.636 16.553 10781.0/6336.1
4th (20.5 km) (A) 1.624 10.588 13.401 20.643 3254.6/2004.9
(B) 1.491 7.402 11.212 17.259 1908.8/1280.5

Please cite this article in press as: M.G. Eltarabily et al., Assessment of slope instability of canal with standard incomat concrete-filled geotextile mattresses lining,
Alexandria Eng. J. (2019), https://doi.org/10.1016/j.aej.2019.11.010
Assessment of slope instability of canal 9

Table 4 Calculations of the percentage of the water depth to the total vertical depth.
Soil type Sec. Bed L.W. Ground- Min. water depth Total ground depth Water depth/total depth Fs
No. level L level (m) (m) (%)
Sandy clay 3-A 7.90 12.40 14.50 4.5 6.6 68.18 2.316
1-B 9.82 14.87 17.00 5.05 7.18 70.33 2.227
1-A 10.50 15.50 18.00 5 7.5 66.67 2.088
3-B 6.51 10.89 15.00 4.38 8.49 51.59 1.760
Fine Sand 4-B 5.40 9.00 10.50 3.6 5.1 70.59 1.983
4-A 5.35 9.42 11.50 4.07 6.15 66.18 1.884
2-A 9.82 14.87 19.00 5.05 9.18 55.01 1.683
Coarse 2-B 7.90 12.40 16.00 4.5 8.1 55.56 2.081
sand

shown in table 3 and compared with the obtained factor of shown in table 3. Also, all factors of safety still higher than
safety for the case of no lining as shown in Fig. 9. The same 1.50 except for the two segments from km 17.5 to km 28 and
mode and mechanism of failures were obtained for each seg- from the km 75 to km 92.5 where the factors of safety were
ment as the previously behaviors which obtained for the case 1.421 and 1.491, respectively.
without the mattresses. It is clear that all the eight segments Additionally, from Table 3, it can be concluded that the slip
attained factors of safety less than the case without the liners. volumes for all eight segments were reduced while installing the
The most affected two sections among the eight segments were lining mattress. Specifically, the reduction in the volume (m3) of
chosen for the results (Fig. 10 a, b) while the other values are the slips in coarse sand and fine sand were higher than the

Fig. 9 Values of the factor of safety for the current statues without lining and after installation of the mattress at the eight segments.

Fig. 10 a, b Factor of safety and the failure surfaces of the most two critical parts among the eight segments after installation of the
40 cm thickness of geotextile concrete-filled mattress.

Please cite this article in press as: M.G. Eltarabily et al., Assessment of slope instability of canal with standard incomat concrete-filled geotextile mattresses lining,
Alexandria Eng. J. (2019), https://doi.org/10.1016/j.aej.2019.11.010
10 M.G. Eltarabily et al.

reduction for the slips for sections in sandy clay soil. The high- Two main sections were chosen among the eight simulated
est reduction of the slip surface was accompanied by the highest sections for expressing the maximum and minimum reduction
reduction in the factor of safety for the most two critical seg- in the factor of safety, (Sec 3-B, and Sec 4-B) where the soil
ments while they were 31.7% and 35.8% for sections 2-A and types are the sandy clay and the fine sand, respectively. The
4-B, respectively. Also, all the radiuses of slip surfaces were critical slip surface of the lowest factor of safety was divided
reduced and the longitudinal length of the arc was decreased into a fixed number of slices, thirty, with ordering starting
as well. While the mattress functionally acts as a liner to control from the slope toe till the end at the top ground surface. For
seepage after filling with impermeable light concrete, its imple- each section, the two cases of no lining and the execution of
mentation is considered an additional surcharge load that the geotextile mattress were compared for the normal stress
changes the boundary conditions which increase the difference (in Kpa) and the pore water pressure (in Kpa) (Fig. 12 a, b).
in the stress levels [33]. The interpretation of the reduced factors It is evident that the pore-water pressure is reduced outwards
of safety is due to the progressive deformations along the sur- the slope until the minimum value was obtained at the top
face of the slope (the part above the water level) or along the of the slope, slice number thirty [35]. The same behavior of
week zone specifically for the coarse soils. Thus, the failures the reduction of the pore-water pressure was obtained for both
propagate downward with the same pattern of the case of no sections while the values are different. At the toe, higher values
lining causing the reduction in radiuses and slip volumes that were recorded for section 3-B than sec 4-B where it is corre-
finally reduce the Factors of safety. sponding to the water depth in the canal. For the last slice at
the top of the slope, lower values of pore-water pressure were
4.3. Factor of safety reduction in relation to soil types observed for sec 3-B than sec 4-B where the depth to the water
table is higher [36–37]. While the slope angle is not the same
After installation of geotextile and filled with concrete, the for both sections, the half-sine curve of the normal stress is
additional load was applied on the slope and the factors of not typical and shifted to the slope toe for the less steep slope,
safety were compared with the case of no lining. Fig. 11 sum- 2:1 than the steeper one, 3:2 where the curve top near to slice
marizes the reductions of factors of safety for the eight seg- fifteen at the middle of the slip surface [38].
ments after lining execution. A significant variance in the After installing the geotextile mattress, an additional verti-
reduction of Fs among the soil types is clear where the maxi- cal force was applied, which leads to the increase in the normal
mum two reductions that equal 21.61% and 15.57% were stress and the obtained reduction of the porewater pressure,
obtained for the lining sections 4-A (from Km 75.0 to km thus; the shear strength function significantly affected. From
92.5) and section 2-A (from km 17.5 to km 28.0) where the soil Fig. 12 a and b, it is clear that the two patterns of the differ-
type is fine sand. Also, for the 10 km length between km 280 to ence in pore-water pressure for the two sections 3-B and 4-B
km 38.0 (lining section 2, segment B), a significant reduction are not the same where the pore water pressure was increased
was obtained, 14.56% where the soil type is coarse sand. On for the first half of the 3:2 slope (till slice number 16) then it
the contrary, for the four segments (1-A, 1-B, 3-A, and 3-B) was decreased for the second half of the slope for the section
where the soil type is sandy clay, the reductions of Fs were 3-B. However, for section 4-B where the soil is fine sand, the
not exceded 5.0%. In order to understand the differences of pore water pressure increased almost along the slip surface till
the soil behavior for the reduction of the factor of safety, the the slice number 28. Additionally, the summations of the thirty
normal stresses at the different plans of failures from the forces differences along the two sections were 31.5 Kpa and 10.7 Kpa
equilibrium diagram were investigated and the pore-water for fine sand and sandy clay soil, respectively. Thus confirm
pressure that governing the drained shear strengths was that; with more increase in pore-water pressure (fine sand
revealed [34]. cases) more reduction in the factor of safety obtained and with

Fig. 11 The reduction in the factors of safety for the eight segments where the significant variance obtained between sandy clay soil and
fine and coarse sand.

Please cite this article in press as: M.G. Eltarabily et al., Assessment of slope instability of canal with standard incomat concrete-filled geotextile mattresses lining,
Alexandria Eng. J. (2019), https://doi.org/10.1016/j.aej.2019.11.010
Assessment of slope instability of canal 11

Fig. 12 The normal stress in Kpa and pore-water pressure without applying any surcharge (no lining) and after applying the surcharge
(geotextile mattress) and the difference in pore-water pressure in Kpa for; (a) the lining section with the minimum reduction of Fs (3-B) (b)
the lining section with the maximum reduction of Fs (4-B) (most significant).

less increase in pore-water pressure, a slight reduction in Fs gies using geosynthetic and geotextile were agreed. The assess-
was recorded (cases of sandy clay sections) [39]. ment of the used lining technique from the geotechnical aspect
in terms of its effect on the canal slope stability could be inte-
5. Conclusion grated with the main project objectives. Due to the heterogene-
ity of the soil lithostratigraphy of the study area, eight
While the seepage losses from irrigation canals are widely con- segments were defined (two for each lining section) and the soil
sidered the main concern for surface water management, type was identified. Other geometrical data, water levels, dis-
numerous efforts were spent from the Ministry of Water charges along the canal were collected in addition to the soil
Resources and Irrigation in Egypt to solve this problem for hydraulic properties and water retention for the sandy clay,
Ismailia canal (the most deteriorate main branch canal from fine sand, and coarse sand. The finite-element GeoStudio pro-
the Nile River). After many feasibilities, and technical studies gram was used to simulate the two-dimensional water flow
and environmental impact assessment reports by the support using the Darcy equation for the unsaturated-saturated condi-
of the African Development Bank, four critical sections of tions using SEEP/W module and the slope stability using Mor-
seepage along the canal were defined and three lining technolo- genstern Price method for equilibrium by SLOPE/W module.

Please cite this article in press as: M.G. Eltarabily et al., Assessment of slope instability of canal with standard incomat concrete-filled geotextile mattresses lining,
Alexandria Eng. J. (2019), https://doi.org/10.1016/j.aej.2019.11.010
12 M.G. Eltarabily et al.

The simulations were run in the transient state of water flow [6] MWRI, Ministry of Water Resources and Irrigation, National
and the water table was calculated thus, the factors of safety Water Resources Plan for Egypt–2017. MWRI, Cairo, Egypt.
for the eight sections were calculated and compared for the January 2005. Available online: http://extwprlegs1.fao.org/docs/
case of the lining with the geotextile incomat concrete-filled pdf/egy147082.pdf
[7] MWRI, Ministry of Water Resources and Irrigation, The Water
mattress with the case of no lining.
Resource Development and Management Strategy in Egypt
The canal slopes at the predefined sections for the proposed until 2050. MWRI, Cairo, Egypt (2010) (in Arabic)
lining are safe for the current case of canal operation with no [8] MWRI, Ministry of Water Resources and Irrigation, Integrated
lining while the minimum factor of safety was 1.68 for the Water Resources Management Plan. MWRI, Cairo, Egypt. June
10.5 km length between the km 17.5 to km 28.0 along the 2005. http://documents.worldbank.org/curated/en/561611468
canal. After implementing the lining, the four segments that 234311417/pdf/341800EGY0whit11public10Action0Plan.pdf
the canal passes through coarse soil (fine sand, coarse sand) [9] AWF, African Water Facility, Comprehensive Study and Project
will be significantly affected where the factors of safety reduced Preparation for the Rehabilitation of the Nubaria and Ismailia
by 15.6% for the 3rd segment (2-A) from km 17.5 to km 28.0, Canals. Appraisal Report, August 2007. Available online: https://
14.6% for the 4th segment (2-B) from km 28.0 to km 38.0, www.africanwaterfacility.org/fileadmin/uploads/awf/Projects/
AWF-Project-appraisal-report-EGYPT-CANALS.pdf
13.8% for the 7th segment (4-A) from km 72.0 to km 75.0,
[10] AFDB, African Development Bank Group, Egypt–Integrated
and 21.6% for the 8th segment (4-B) from km 75.0 to km Water Resources Management of the Nubaria and Ismailia
92.0. However, among those four segments, only two of them Canals – Summary ESIA, Environmental and Social Impact
attained factors of safety less than 1.50. These sections are the Assessment. May 2016. Project code: P-EG-AAC-024. Available
10.5 km length between km 17.5 and km 28.0 and the last lin- online: https://www.afdb.org/en/documents/document/egypt-
ing segment, 17.50 km length between the km 75.0 and km integrated-water-resources-management-of-the-nubaria-and-
92.5. While the reduction of factors of safety for the sandy clay ismailia-canals-summary-esia-88494
soils is insignificant, the canal slopes still safe with no failure [11] GeoStudio, Tutorial Manual, GEO-SLOPE International Ltd,
expect to happen even during the construction or after the (2004). www.geo-slope.com.
implementation during the canal operation under lining condi- [12] H. Anderson, P-E. Bengtsson, C. Berglund, C. Larsson, R.
Larsson, G. Sallfors, A-L. Oberg-Hogsta, The landslide at
tions. It is highly required to take precautions during the lining
Vagnharad in Sweden. In: Bromhead E, Dixon N, Ibsen M-L,
installation for the concrete filling process to ensure the unifor- editors. Landslides in research, theory, and practice. Eighth
mity of the 40 cm thickness of the mattresses. International Symposium on Landslides (2000) London: T.
Telford; p. 65_70.
Declaration of Competing Interest [13] R. Schnellmann, M. Busslinger, H.R. Schneider, H. Rahardjo,
Effect of rising water table in an unsaturated slope, Eng. Geol.
The authors declare that they have no known competing 114 (1–2) (2010) 71–83, https://doi.org/10.1016/j.enggeo.
financial interests or personal relationships that could have 2010.04.005.
appeared to influence the work reported in this paper. [14] K. Acharya, N. Bhandary, R. Dahal, R. Yatabe, Numerical
analysis on influence of principal parameters of topography on
hill slope instability in a small catchment, Environ. Earth Sci. 73
Acknowledgments (9) (2015) 5643–5656, https://doi.org/10.1007/s12665-014-3819-z.
[15] N.R. Morgenstern, V.E. Price, The analysis of the stability of
The corresponding author would like to thank Fulbright pro- general slip surfaces, Geotechnique 15 (1) (1956) 79–93, https://
gram for their financial support during his post-doctoral schol- doi.org/10.1680/geot.1965.15.1.79.
arship at University of California, USA. Also, we would like to [16] M.O. Arnous, D.R. Green, Monitoring and assessing
thank the journal editor and the reviewers for their waterlogged and salt-affected areas in the eastern Nile Delta
consideration. region, Egypt using remotely sensed multi-temporal data and
GIS, J. Coastal Conserv. 19 (2015) 369–391, https://doi.org/
10.1007/s11852-015-0397-5.
References [17] M.G. Eltarabily, A.M. Negm, C. Yoshimura, O.C. Saavedra,
Modeling the impact of nitrate fertilizers on groundwater
[1] FAO, Food and Agriculture Organization of the United quality in the southern part of the Nile Delta, Egypt, Water
Nations. Main problems in a canal network, Chapter 4 (2018). Sci. Technol. – Water Supply 17 (2) (2016) 561–570, https://doi.
http://www.fao.org/3/ai585e/ai585e02.pdf org/10.2166/ws.2016.162.
[2] Q. Zhang, J. Chai, Z. Xu, Y. Qin, Investigation of irrigation [18] M.H. Geriesh, K. Balke, A.E. El-Rayes, B.M. Mansour,
canal seepage losses through use of four different methods in Implications of climate change on the groundwater flow
Hetao irrigation district, China, J. Hydrol. Eng. 22 (3) (2016), regime and geochemistry of the Nile Delta, Egypt, J. Coastal
https://doi.org/10.1061/(ASCE)HE.1943-5584.0001470. Conserv. 19 (2015) 589–608, https://doi.org/10.1007/s11852-015-
[3] Z. Huang, Y. Bai, H. Xu, Y. Cao, X. Hu, A theoretical model to 0409-5.
predict the critical hydraulic gradient for soil particle movement [19] A.M.A. Ismael, Application of remote sensing, GIS, and
under two-dimensional seepage flow, Water 9 (11) (2017) 828, groundwater flow modeling in evaluating groundwater
https://doi.org/10.3390/w9110828. resources: two case studies—East Nile Delta, Egypt, and Gold
[4] X. Xie, J. Liu, B. Han, H. Li, Y. Li, X. Li, Critical hydraulic Valley, California (Unpublished doctoral dissertation) (2007).
gradient of internal erosion at the soil-structure interface, University of Texas, El-Paso, USA
Processes 6 (7) (2018) 92, https://doi.org/10.3390/pr6070092. [20] M.H. Mowafy, Seepage losses in Ismailia canal. 6th International
[5] E. Blond, S. Boyle, M. Ferrara, B. Herlin, H. Plusquellec, P. Water Technology Conference (IWTC), Alexandria, Egypt
Rimoldi, S. Timothy, Applications of geosynthetics to irrigation, (2001). http://www.iwtc.info/2001_pdf/03-3.pdf
drainage, and agriculture, Irrig. Drain. 19 (2018), https://doi. [21] M.T. Van Genuchten, A closed-form equation for predicting
org/10.1002/ird.2300. the hydraulic conductivity of unsaturated soils, Soil Sci. Soc.

Please cite this article in press as: M.G. Eltarabily et al., Assessment of slope instability of canal with standard incomat concrete-filled geotextile mattresses lining,
Alexandria Eng. J. (2019), https://doi.org/10.1016/j.aej.2019.11.010
Assessment of slope instability of canal 13

Am. J. 44 (1980) 892–898, https://doi.org/10.2136/sssaj1980. element strength reduction techniques. Part I: Numerical studies
03615995004400050002x. considering non-associated plasticity., Comput. Geotech. 70
[22] M.G. Eltarabily, A.M. Negm, C. Yoshimura, J. Takemura, (2015) 169–177, https://doi.org/10.1016/j.compgeo.2015.06.018.
Groundwater modeling in agricultural watershed under different [32] N. Lu, J. Godt, Infinite slope stability under steady unsaturated
recharge and discharge scenarios for quaternary aquifer Eastern seepage conditions, Water Resour. Res. 44 (2008) W11404,
Nile Delta, Egypt, Environ. Model. Assessm. 23 (3) (2018) 289, https://doi.org/10.1029/2008WR006976.
https://doi.org/10.1007/s10666-017-9577-z. [33] K. Lendering, T. Schweckendiek, M. Kok, Quantifying the
[23] D.G. Fredlund, N.R. Morgenstern, A. Widger, The shear failure probability of a canal levee, Georisk Assessm. Manage.
strength of unsaturated soils, Can. Geotech. J. 15 (1978) 313– Risk Eng. Syst. Geohazards 12 (3) (2018), https://doi.org/
321, https://doi.org/10.1139/t78-029. 10.1080/17499518.2018.1426865.
[24] D.G. Fredlund, H. Rahardjo, Soil mechanics for unsaturated [34] J.-C. Cai, T.-C.J. Yeh, E.-C. Yan, R.-X. Tang, Y.-H. Hao, S.-Y.
soils. New York (NY) (1993): John Wiley & Sons Huang, J.-C. Wen, Importance of variability in initial soil
[25] N. Sabhahit, A.U. Rao, Genetic algorithms in stability of non- moisture and rainfalls on slope stability, J. Hydrol. 571 (2019)
homogeneous slopes, Int. J. Geotech. Eng. 5 (1) (2011) 33–44, 265–278, https://doi.org/10.1016/j.jhydrol.2019.01.046.
https://doi.org/10.3328/IJGE.2011.05.01.33-44. [35] A. Perrone, R. Vassallo, V. Lapenna, C. Di Maio, Pore water
[26] K. Gavin, J.F. Xue, Use of a genetic algorithm to perform pressures and slope stability: a joint geophysical and
reliability analysis of unsaturated soil slopes, Geotechnique 59 geotechnical analysis, J. Geophys. Eng. 5 (2008) 323–337,
(6) (2009) 545–549, https://doi.org/10.1680/geot.8.T.004. https://doi.org/10.1088/1742-2132/5/3/008.
[27] W.J. Niu, Determination of slope safety factor with analytical [36] B. Kumar, G. Sreenivasulu, A. Ramakrishna Rao, Metamodel-
solution and searching critical slip surface with a genetic- based design of alluvial channels at incipient motion subjected to
traversal random method, Sci. World J. (2014), https://doi.org/ seepage, Hydrol. Sci. J. 55 (3) (2010) 459–466, https://doi.org/
10.1155/2014/950531. 10.1080/02626661003741542.
[28] GeoStudio, Stability Modeling with SLOPW/E. An Engineering [37] B. Yang, K. Yin, T. Xiao, C. Lixia, J. Du, Annual variation of
Methodology. GEO-SLOPE International Ltd (2012), landslide stability under the effect of water level fluctuation and
www.geo-slope.com. rainfall in the Three Gorges Reservoir, China, Environ. Earth
[29] W. Fu, Y. Liao, Non-linear shear strength reduction technique Sci. 76 (2017) 564, https://doi.org/10.1007/s12665-017-6898-9.
in slope stability calculation, Comput. Geotech. 37 (3) (2010) [38] K.P. Aryal, Slope stability evaluations by limit equilibrium and
288–298, https://doi.org/10.1016/j.compgeo.2009.11.002. finite element methods. Ph.D. Thesis, Norwegian University of
[30] A. Gholami, H. Bonakdari, M. Mohammadian, A.H. Zaji, B. Science and Technology, Trondheim, Norway (2006).
Gharabaghi, Assessment of geomorphological bank evolution of https://core.ac.uk/download/pdf/30817581.pdf
the alluvial threshold rivers based on entropy concept [39] W. Jian, Q. Xu, H. Yang, F. Wang, Mechanism and failure
parameters, Hydrol. Sci. J. 64 (7) (2019) 856–872, https://doi. process of Qianjiangping landslide in the Three Gorges
org/10.1080/02626667.2019.1608995. Reservoir, China, Environ. Earth Sci. 72 (8) (2014) 2999–3013,
[31] F. Tschuchnigg, H.F. Schweiger, S.W. Sloan, Slope stability https://doi.org/10.1007/s12665-014-3205-x.
analysis by means of finite element limit analysis and finite

Please cite this article in press as: M.G. Eltarabily et al., Assessment of slope instability of canal with standard incomat concrete-filled geotextile mattresses lining,
Alexandria Eng. J. (2019), https://doi.org/10.1016/j.aej.2019.11.010