Content uploaded by Saadet Arzu Berilgen
Author content
All content in this area was uploaded by Saadet Arzu Berilgen on Nov 21, 2019
Content may be subject to copyright.
8th Geotechnical Symposium
13- 15 November 2019, İTÜ Süleyman Demirel Kültür Merkezi, Istanbul
CHARACTERISTICS AND FAILURE MECHANISM OF A
HIGHWAY LANDSLIDE
BİR KARAYOLU HEYELANININ KARAKTERİSTİKLERİ VE GÖÇME
MEKANİZMASI
Murat Tonaroğlu*
1
Murat E. Selçuk
2
Saadet A. Berilgen
3
Mustafa Yıldırım
4
ABSTRACT
The construction of a divided state highway in Kemerburgaz, Istanbul began in 2008.
In 2009, a landslide occurred on a slope with an inclination of 10%, situated in a
section of Km 16+599.338 to Km 17+170.770 of the road route. In accordance with the
results from various studies conducted following the occurrence of this landslide, a 570
m long double-row retaining structure composed of several bored piles was built in
order to stop the sliding and to prevent other potential mass movements.
However, in spite of such a precaution, the land started to slide again in December
2010 and destroyed the huge retaining structure composed of piles. For such
problematic areas, it is essential that geological and geotechnical studies be conducted
in more detail before the beginning of construction. This study investigates the causes
of the landslide and conducts a detailed analysis on why a retaining structure
composed of piles was inadequate to stop the ongoing landslide or to prevent potential
landslides in the area.
ÖZET
İstanbul Kemerburgaz’da, 2008 yılında, bölünmüş bir yol inşaası başlamıştır. 2009 yılında,
yol güzergahı üzerinde, 16+599.338 ile 17+170.770 Km’ler arasında yaklaşık %10 eğimli
bir kesimde yer alan şevde bir heyelan meydana gelmiştir. Bu heyelanın meydana
gelmesinin ardından yürütülen çeşitli araştırmaların sonuçlarına bakılarak, mevcut şev
kaymasını durdurmak ve olası başka zemin problemlerini önlemek amacıyla yaklaşık 570
m uzunluklu çift sıra kazıktan oluşan bir istinat yapısı inşa edilmiştir. Buna karşılık, böyle
bir önlem alınmasına rağmen 2010 yılının aralık ayında hareket yeniden başlamış ve inşa
edilmiş istinat yapısı yıkılmıştır. Bu tür problemli sahalarda, inşa başlamadan evvel detaylı
bir geoteknik ve jeolojik inceleme yapılması zorunludur. Bu bildirioluşan heyelanın
nedenleri ve inşa edilmiş devasa istinat yapısının hareketi durdurmak ve önlemekte neden
yetersiz kaldığını incelemektedir.
*
1
Asst. Prof., Yıldız Technical University Civil Engineering Faculty, Geotechnics Division , tonar@yildiz.edu.tr
2
Asst. Prof., Yıldız Technical University Civil Engineering Faculty, Geotechnics Division, meselcuk@yildiz.edu.tr
3
Assoc. Prof., Yıldız Technical University Civil Engineering Faculty, Geotechnics Division, koc@yildiz.edu.tr
4
. Prof., Yıldız Technical University Civil Engineering Faculty, Geotechnics Division, yildir@yildiz.edu.tr
Murat Tonaroğlu, Murat Ergenokon Selçuk, Saadet Arzu Berilgen, Mustafa Yıldırım
1. INTRODUCTION
In 2008, excavations began for the construction of a divided state road in Kemerburgaz
district 40 of Istanbul, Turkey. During the construction process in 2009, a large-scale
landslide occurred on a slope with an average inclination of 10%, situated in a section Km
16+599.338 to Km 17+170.770 of the highway route. The area of the landslide was rich in
lignite coal seams and was mined extensively from 1915 to 2000s; during this period,
approximately 80 meter-deep excavations were made. A few of these coalmines are still in
operation. In the past, theexcavation debris was generally scattered to the neighboring
areas leading to the formation ofartificial fills and low-laying areas at the size of artificial
hills causing the area to be abandoned. In some low-laying areas in the region, which have
also been chosen as the construction area for the third airport in Istanbul, the surface area
reaches up to 100-200 acres and artificial lakes have been formed due to the surface water.
In the last few years, some of these lakes to the south of the investigated area were
declared by the municipality to be dumping sites for the excavation debris, and therefore
they have been completely filled to create vast plains.
Within the scope of this study, several boreholes with a depth of 25.0 m were drilled along
the section of the landslide, and inclinometers were placed in these boreholes for soil
analysis. Furthermore, in May 2010, a further 13 boreholes with a depth ranging from 17
to 30 m were drilled, and inclinometers were placed in 8 boreholes. In this way, the
inclinometer measurements were used to determine the depth of the sliding surfaces as
well as the amount and direction of ground motions. As a result, the data that was obtained
indicated that the main sliding plane was 13.0 m deep, and a piled structure was
constructed in November 2010 to increase the stability of the slope and prevent future
landslides. However, the landslide continued.
2. MORPHOLOGICAL AND GEOLOGICAL STRUCTURE OF THE
INVESTIGATION AREA
The morphological structure of the investigation area is comprised of; slopes that are not
very steep with an overall height of less than 150 m, lakes formed in dry riverbeds, and pits
between these slopes. This morphological structure was formed as a result of the cemented
clay and sandy sediments in the region and the intercalation of the clay-sand sediments
deposited in the Upper Oligocene–Upper Miocene Age. These sediments which were
formerly known as Karaburun Formation, Ağaçlı Formation, and Gürpınar Formation,
were later collectively called Avcılar Formation; and mapped as a regular formation as
seen in Figure 1. Starting with a terrigenous character, this formation moves to the shallow
marine environment, intercalated bylayers of clay and sand including coal layers [1, 2, 3, 3,
4, and 5]
Avcılar formation is a regular sequence that starts from Istanbul and its surroundings,
outcropping along the districts of Avcılar-Büyükçekmece, intercalated by highly fissured
over-consolidated clay and sand layers [5]. This sequence is also widely observed in the
Kemerburgaz district along the Black Sea coastline on the Asian side.
Characteristics and formation mechanism of a highway landslide
Figure 1. General geological map of Istanbul and its surrounding area [5]
The ground levels of the Avcılar formation either start with gravelly sand or black colored
clay and the coal levels have a high organic content. As for higher levels, the dominant
lithology of the sequence is constituted by green clay consisting of brownish gray sand
lenses, and over- consolidated clay and sand interlayers as seen in Figure 2.
Throughout the whole investigation area and the surrounding areas, the dominant lithology
is that of the Avcılar formation. The lithology named as the Bakırköy formation, which is
composed of marl-limestone that constitute the highest levels of the Avcılar formation is
not observed in Kemerburgaz district. The over-consolidated clay level of the Avcılar
formation can be categorized into the high-plasticity clay group, which has a stiff-very
stiff-hard soil character. Except for the sand lenses, the sequence does not contain
underground water and the intercalation of clay-sand layers may have a very weak rock
character such as claystone interlayers or rarely a medium hard rock character such as gray
sandstone interlayers in certain regions. At different levels of the Avcılar formation, gray-
brown sand, gravelly sand or sand with white mica and gravel are frequently observed
between clays as interlayers with a thickness of 10-30 m. Sand that is generally quartz-rich,
uncemented or loose-cemented is poorly graded; silty, clayey sands and gravel layers are
classified as clayey gravel, and these sand layers include the interbeds of over-consolidated
clay that are 10 to 30 cm thick. The lithology of the Avcılar formation also includes
blackish-dark gray clays of less than 10 m with a high organic content, swelling potential
and plasticity at a rare frequency and different levels, and their lateral continuity is
limited. These clay lenses, due to their high organic content, are significant in terms
of engineering properties and may cause severe stability problems particularly in the
inclined areas. Furthermore, the analysis of many boreholes in the area showed that the soil
profile was composed of over-consolidated clayey soil layers including partly sand bands
underlying artificial fills that are a few meters thick below the surface as the main
formation of the region.
Murat Tonaroğlu, Murat Ergenokon Selçuk, Saadet Arzu Berilgen, Mustafa Yıldırım
Figure 2. The aspect of updated stratigraphy and typically observed deposits [5]
3. STUDIES PERFORMED IN THE SITE UNDER INVESTIGATION
As mentioned above, following a huge landslide that occurred in the investigation area, a
piled- retaining structure was built in November 2010 as a precaution to protect the site
from the possibility of instability. This structure was located adjacent to the north of the
road and parallel to the road that lay in the east west direction and was 570 m long. The
structure consisted of 3.6 m spaced double-row bored piles which were minimum 15.5 m
and maximum 28.0 m long with a diameter of 120 cm, and a connecting frame of 5.20 m.
However, as seen in Figure 3 a and b, this massive retaining structure failed to stop the
movement and got largely destroyed as seen in Figure 4.
Characteristics and formation mechanism of a highway landslide
a) 14 October 2011
b) December 2013
Figure 3. The ongoing movement of the active landslide.
Figure 4. The piled-retaining structure damaged by the landslide
In addition to the boreholes drilled during the investigation of previous studies conducted
by various companies in the area, new and much deeper boreholes were planned to be
drilled for the execution of the current study. As a result, 15 new boreholes reaching the
depth of 40 to 60 m were drilled particularly in the locations where the landslide is very
active. In addition, new inclinometers were placed in six of these boreholes as presented in
Figure 5. The cross section of the landslide is illustrated in Figure 6.
Murat Tonaroğlu, Murat Ergenokon Selçuk, Saadet Arzu Berilgen, Mustafa Yıldırım
Figure 5. The map of the landslide and the depths/locations of boreholes in Kemerburgaz
The Table 1 presents the measurement results obtained from the inclinometers installed in
the boreholes until January 2012. The results of these measurements indicated that the
main sliding surface was deeper (reaching 35.0-40.0 m) than the estimated slip surface (10-
12 m) which was used to design the length of the piles for the retaining structure.
Table 1. Inclinometer results
The results from the YSK-2 borehole showed that the movement was ≥ 35.0 m and the
main sliding direction of this movement was 33 mm to the north and 26 mm to the east. On
7th February 2012, the inclinometer was broken by the main sliding plane at 35.0 m depth
(Figure 7). The results from the YSK-4 borehole showed that the movements occurred
more frequently in sections closer to the surface and slightly less in deeper sections.
According to the final reading, the value of the movement was 18 mm to the north and 7
mm to the east (Figure 8).The results from the YSK-5 borehole indicated that lateral
displacements occurred in sections closer to the surface and the final readings revealed that
Characteristics and formation mechanism of a highway landslide
the movement was 8 mm to the north and 15 mm to the west. However, these findings do
not provide a definitive evidence for the occurrence of a sliding surface (Figure 9).
For the YSK-7 borehole, the lateral displacements were mainly found at least 36.0-37.0 m
down from the surface. The reading dated 16th December 2011 showed that the movement
was 125 mm to the north and 24 mm to the west. In addition, since the inclinometer pipe
was found be cut by the slip surface at 36.0 m, the following readings were only taken at
36 m and above (Figure 10). The results from the YSK-10 borehole showed that the
sliding depth was 21.0 m and the inclinometer reading dated 20th January 2012 revealed
that a displacement occurred 125 mm to the north and 115 mm to the west. On 7th
February 2012, the inclinometer pipe was completely broken at 21.0 m below the ground
surface. Considering the data obtained from other close boreholes, the main sliding surface
here was most probably at a depth of around 35.0-36.0 m (Figure 11).
Figure 6. The cross-section of the landslide
Murat Tonaroğlu, Murat Ergenokon Selçuk, Saadet Arzu Berilgen, Mustafa Yıldırım
Figure 7. Inclinometer measurements for the YSK2 borehole
Figure 8. Inclinometer measurements of the YSK4 borehole in Kemerburgaz.
Characteristics and formation mechanism of a highway landslide
Figure 9. Inclinometer measurements of the YSK5 borehole in Kemerburgaz
Figure 10. Inclinometer measurements at YSK7 borehole in Kemerburgaz
Murat Tonaroğlu, Murat Ergenokon Selçuk, Saadet Arzu Berilgen, Mustafa Yıldırım
Figure 11. Inclinometer measurements taken from the YSK10 borehole in Kemerburgaz
4. CAUSES OF THE LANDSLIDE
This study revealed the possible landslide mechanism of the investigated area through a
combined evaluation of borehole data and inclinometer measurements as well as mapping
detailed sections of the area and marking the sliding surfaces. As a result of the stability
analyses conducted using residual soil parameters that are considered valid for the
landslide conditions, it was concluded that the major factor causing the occurrence and
continuity of the huge landslides was the lack of natural and/or artificial drainage systems
that would have removed the surface water. On the top section of the southern slope along
the road where the landslide occurred a large surface is covered by low-laying areas that
were previously used as coal mines. Overlying highly impermeable clay layers, these low-
laying areas had been filled with the surface water for years and were transformed into
artificial lakes in the region. Furthermore, the data obtained from the large number of
boreholes drilled in the area shows that there were sand bands continuing down the slope
which are related with the water-level of the artificial lakes in the hillside of the region and
this most probably increased the pore water pressure and the hydraulic gradient leading to
water flow. In recent years, these low laying areas have been used for excavation wastes
and some have been completely filled.
The landfills that exceeded the water-level of the artificial lakes generated pressure on the
water mass and caused the water to seep into the sand layers on the slope towards the
north. Accordingly, the water seeping from the layers of sand reduced the shear strength
of the clay units and formed slide planes.. The surface water still continues to seep into the
ground. Furthermore, the dynamic effect of high-tonnage excavation trucks (reaching a
number of 1200 shipments on a daily basis) moving to the dump area have serious negative
effects on the stability of the slope. In addition to the destruction of natural drainage
Characteristics and formation mechanism of a highway landslide
channels as a result of the landslide, there is still a lack of an effective drainage system that
could remove the surface water in a controlled manner.
5. STABILITY ANALYSIS
After the soil layers of the investigated area were examined and the causes of the landslide
were determined, stability analysis was conducted using back analysis to determine the soil
strength parameters that are primarily valid for the field conditions. The TALREN
software was used for all the analyses and the SPT N values were obtained from the large
number of boreholes drilled in the area. The main deep sliding surface was determined
using the limit equilibrium analysis and inclinometer measurements. Using the data
obtained from inclinometers the parameters which generated a factor of safety of 1.0 along
the slip surface were determined for the stability analysis (Figure 12).
Figure 12. Numerical analysis results for FS=1.0
In addition, ’r =8⁰-12⁰ range and c’r =0 value were used as the required parameters of
residual shear strength for similar layers, and the excess pore water pressure coefficient
value (ru) was obtained for FS=1.0. As a result of the numerical analysis, the parameters
for residual resistance to sliding with the highest possible validity for landslide areas and
water pore-pressure coefficient value were found to be as follows; r =10⁰,c’r = 0 ,ru= 0.26
With the same parameters for residual resistance to sliding, FS=1.23 is calculated as ru =
0.10 and FS=1.30 is calculated as ru = 0.05. The results of the numerical analysis indicate
that adequate safety against sliding can be ensured on the condition that sufficient drainage
is provided and the occurrence of water pore-pressure on the slope is prevented. To
achieve this, the leakage from old artificial lakes located on the top section of the southern
slope along the road should be prevented and an effective drainage system should be
established for the surface water.
Murat Tonaroğlu, Murat Ergenokon Selçuk, Saadet Arzu Berilgen, Mustafa Yıldırım
6.CONCLUSIONS
This paper reports on a study that investigated the causes and the prevention methods of
the landslide that occurred between Km 16+599.338-Km 17+170.770 of the road route
during the construction of a state highway in Istanbul which is an ongoing process. To this
end, in addition to the data obtained from the reports of previous research conducted in the
area, the data obtained from the new boreholes drilled in the investigation area and the
inclinometers installed in the boreholes showed that the continual landslide in the field
largely occurred due to the lack of a sufficient drainage system for the surface water and
the decrease in the shear strength increased the pore water pressure on the clay units
leading to the formation of sliding planes.
It was also found that the reason for the failure of the piled structure that was built to stop
and/or prevent the repetition of the landslide was the landslide being much deeper (35-40
m) than the estimated value (10-12 m). Therefore, the large excavated area where the third
İstanbul Airport is currently being constructed has recently been further excavated below
the elevation of the main sliding plane to ensure the safety of the construction.
REFERENCES
[1] Yıldırım M (1997) Engineering geological evaluation of solid waste landfill sites: two
examples from Istanbul, Turkey. Bull Eng Geol 55:151–158.
[2] Yıldırım M (2000) Engineering properties of Istanbul area rocks. Journal of Yıldız
Technical University,Istanbul 3:9–23.
[3] Yıldırım M, Savaşkan E. (2002) The stratigraphy and engineering properties of Istanbul
Tertiary deposits. Bull Turkish Natl Comm Int Assoc Eng Geol 18:48–62.
[4] Yıldırım,M., Akgüner, C., Tonaroğlu, M., Selçuk, M.E., (2010) İstanbul Tersiyer
Çökellerinin Yeni Stratigrafisi ve Mühendislik Özellikleri, Sigma Mühendislik ve
Fen Bilimleri Dergisi, Vol. 4, Issue 28, pp. 323-334.
[5] Yıldırım,M., Tonaroğlu, M., Selçuk, M.E., Akgüner, C., (2013) Revised Stratigraphy of
the Tertiary Deposits of Istanbul and Their Engineering Properties, Bulletin of
Engineering Geology and the Environment, Volume 72, pp. 413-420.