Soils and Foundations 2015;55(3):600–613

HOSTED BY The Japanese Geotechnical Society

Soils and Foundations

www.sciencedirect.com
journal homepage: www.elsevier.com/locate/sandf

Mineralogy and geotechnical properties of Singapore marine clay at Changi

Myint Win Boa, Arul Arulrajahb, Patimapon Sukmakc, Suksun Horpibulsukd,n
a
DST Consulting Engineers Inc., Thunder Bay, Ontario, Canada
b
Swinburne University of Technology, Melbourne, Australia
c
School of Engineering and Resources, Walailak University, Nakhonsithammarat, Thailand
d
School of Civil Engineering, Suranaree University of Technology, 111 University Avenue, Muang District, Nakhon Ratchasima 30000, Thailand
Received 3 June 2014; received in revised form 8 January 2015; accepted 2 February 2015
Available online 8 May 2015

Abstract

An engineering geological study was undertaken to determine the geotechnical properties and mineralogy of Singapore marine clay at Changi
in the Republic of Singapore. This soft soil is a quartenary deposit that lies within submarine valleys cut in an old alluvium formation. The marine
clay comprises a soft upper marine clay layer overlying a stiffer lower marine clay layer. An intermediate stiff clay layer is sandwiched between
these two marine clay layers; it is believed to be the dessicated crust of the lower marine clay layer. In the present study, morphological and
mineralogical observations of Singapore clay were taken by scanning electron microscopy, X-ray diffraction and photographic identification. The
geotechnical investigation included physical, compression, permeability and field vane shear tests. The upper marine clay was found to be soft
with undrained shear strength values ranging from 10 to 30 kPa, while the lower marine clay was found to have undrained shear strength values
from 30 to 60 kPa. The sensitivity of the marine clay at Changi varied from 3 to 8 and is described as highly sensitive marine clay. The upper clay
layer had a coefficient of consolidation of 0.47–0.6 m2/year due to the vertical flow (cv) and a coefficient of consolidation of 2–3 m2/year due to
the horizontal flow (ch). The lower marine clay had a cv of 0.8–1.5 m2/year and a ch of 3–5 m2/year, while the intermediate stiff clay had a cv of
1–4.5 m2/year and a ch of 5–10 m2/year. The primary clay mineral was kaolinite, followed by smectite and mica. With this high kaolinite content,
the activity of the Singapore marine clay was found to be low at approximately 0.5–1.3 and classified as inactive to active clay. With reference to
the intrinsic state line, natural Singapore clay (upper, intermediate and lower layers) has been classified as structured clay and is in a meta-stable
state. The compression behaviour can be modelled by the Intrinsic State Line and the destructuring framework. The predicted and the measured
compression curves were comparable for various depths. The successful modelling of compression curves will be useful for settlement
calculations and for further research on the development of a constitutive model for Singapore marine clay.
& 2015 The Japanese Geotechnical Society. Production and hosting by Elsevier B.V. All rights reserved.

Keywords: Marine clay; Soft soils; Field vane; Geotechnical; Geology; Hydraulic; Compressibility

1. Introduction

Over the last two decades, extensive land reclamation and

n
Corresponding author.
E-mail addresses: mwinbo@dstgroup.com (M.W. Bo),
aarulrajah@swin.edu.au (A. Arulrajah), patimapon.su@wu.ac.th (P. Sukmak),
suksun@g.sut.ac.th (S. Horpibulsuk).
Peer review under responsibility of The Japanese Geotechnical Society.
ground-improvement projects have been undertaken as part of
the mega Changi Land Reclamation project. As part of these
projects in the Republic of Singapore, the geotechnical in-situ
testing (Bo et al., 2000, 2003, 2013; Chu et al., 2002; Arulrajah
et al., 2004b, 2005, 2006a, 2006b, 2007, 2009a, 2011) and the
field instrumentation (Arulrajah et al., 2004a, 2009b) of these

http://dx.doi.org/10.1016/j.sandf.2015.04.011
0038-0806/& 2015 The Japanese Geotechnical Society. Production and hosting by Elsevier B.V. All rights reserved.
 M.W. Bo et al. / Soils and Foundations 55 (2015) 600–613
geotechnical projects have been reported in recent years.
However, there is still a lack of understanding of the engineer-
ing geology, the mineralogical properties and the compressi-
bility of this particular marine clay, as these aspects have not
been previously reported. In various other countries, detailed
geological and geotechnical engineering characterisations of
marine clays have been undertaken, but they are still lacking
for Singapore marine clay.
Karakouzian et al. (2003) reported on shear strength mea-
surements taken in Norfolk, USA with the usage of field vane
shear tests. Jung et al. (2012) reported on stress–strain responses
of reconstituted glacial clays in Chicago, USA. Horpibulsuk
et al. (2007) and Abuel-Naga et al. (2009) reported on the
engineering properties and geological history of soft Bangkok
clay, Thailand. The mineralogical, chemical and geotechnical
properties of Bangkok clay were compared with those of Ariake
clay, a Japanese clay, by Ohtsubo et al. (2000). Kim et al.
(2012) reported on the engineering properties and geological
history of soft clays in Songdo, South Korea. Yan and Ma
(2010) characterised marine deposits in Macau. Liu et al. (2011)
characterised marine clays in Lianyougang. The results of these
geological studies, which included the mineralogical and
geotechnical investigation of shear strength, hydraulic conduc-
tivity and compression, are vital for research and design. An
understanding of the compressive behaviour is essential for
geological and geotechnical engineering purposes and is the
core basis for modelling the stress–strain relationships of soils
(e.g., Pestana and Whittle, 1995; Hong and Onitsuka, 1998;
Potts and Zdravkovic, 1999; Baudet and Stallebrass, 2001; Chai
et al., 2004; Liu et al., 2011, 2013).
Along the same lines, authors' intention in this study is to
report on the geological aspects of Singapore marine clay at
Changi inclusive of the geological history, clay mineralogy
and geotechnical properties such as physical, compression,
hydraulic conductivity and shear strength. The research site is
located in the Eastern part of the Republic of Singapore where
Singapore marine clay is prevalent. This soft clay is a
quartenary deposit that lies within submarine valleys cut in
an old alluvium formation and is locally known as the Kallang
formation. The study area was submerged underwater with a
seabed elevation varying from
2 mCD to
8 mCD (Admir-
alty Chart Datum, where the mean sea level is þ 1.6 mCD) and
is underlain by marine clay up to 40 m in thickness in certain
areas. Detailed geotechnical properties and a mineralogical
analysis of the Singapore marine clay were undertaken in this
study. The in-situ stress state of Singapore marine clay was
examined and the compression behaviour was analysed. The
destructuring framework, which was proposed by Liu and
Carter (1999, 2000) and verified by some soils, was adapted to
model the compression curves of natural Singapore marine
clay. The predicted and measured compression curves were
compared to illustrate the applicability of the framework.
2. Geological history
The marine clay identified in this study belongs to the
marine member of the Kallang Formation (Public Works
601
Department Singapore, 1976). This formation is underlain by
old alluvium. The preliminary site investigation and geophy-
sical survey of the project site revealed that the Singapore
marine clay at Changi consists of two marine members locally
known as the upper and lower marine clay layers. These soft to
medium stiff clay members are recent deposits of estuarine
origin. The upper and lower marine clays are separated by a
layer of medium stiff to stiff clay with a thickness of 2–5 m.
This layer, locally termed as intermediate clay, is reddish in
colour and is believed to be the desiccated crust of the lower
marine clay resulting from the exposure of the seabed to the
atmosphere during the rise and fall of the sea levels in the
geological past.
The onset of the Wurm/Wisconsin glacial period, occurring
approximately 75,000 years ago, brought about an extremely
rapid drop in sea level to about 140 m below the present sea
level approximately 18,000 years ago. There has been a
tremendous transgression of the sea over the land in the last
10,000–20,000 years. With the further uplift of land and the
regression of the seas, more erosion and deposition have taken
place. The present marine clay formation in Eastern Singapore
has been formed by these cycles of aggregation and erosion
throughout the geological ages (Public Works Department
Singapore, 1976). This conclusion is based on evidence that
desiccation occurred on the intermediate clay layer during the
recession of the sea level after the ice age.
3. Site and laboratory investigations
3.1. Site investigation
Marine bathymetric surveys, together with marine geophy-
sical seismic reflection surveys of the project area, were carried
out with the help of a water surface-towed boomer profiling
system. The elevations of the bases of the compressible layers
and the distribution of soft marine clay pockets deposited in
submarine valley cuts were determined from marine bathy-
metric and seismic reflection surveys.
The seismic survey vessel was fitted with a trisponder, an
echo sounder and a boomer profiling system. The survey
operation was run at lines with spacing of 50 m in alternate
directions. Cross lines were run at a spacing of 50 m in
alternate directions. For horizontal control of the survey, a
trisponder positioning system was used to control the location
of the survey vessel along the pre-computed lines. For vertical
control of the survey, the tidal reduction of the survey area was
carried out using tides observed with a tide gauge. Tidal data
was obtained from the Hydrographical Department of the Port
of Singapore.
The echo sounder enabled the contouring of the seabed
elevation profile, while the boomer enabled the plotting of the
isoline of the base elevation of the marine clay deposit.
A description of the geological sequence was then obtained
based on the interpretation of the boomer data, and this could
be correlated with the marine borehole data provided over the
602 M.W. Bo et al. / Soils and Foundations 55 (2015) 600–613
Fig. 1. Typical isoline showing total thickness of marine clay.
survey area. Fig. 1 shows a typical isoline and presents the
total thickness of marine clay in the study area.
Site investigation works, consisting of marine boreholes,
field vane tests and laboratory tests, were carried out for this
study. Up to 50 marine sampling boreholes were prepared.
Undisturbed samples taken from the boreholes were tested in
the laboratory to determine the physical, mineralogical,
hydraulic conductivity and compression characteristics of the
Singapore marine clay. Undrained shear strengths were mea-
sured using field vane shear tests. A penetration type of vane
shear test, without a borehole, was used in this study. The vane
used here was 13 cm in height and 6.5 cm in diameter. Sleeved
torque rods were used to eliminate friction between the vane
rod and the subsoils. The vane was withdrawn into a protective
shoe to avoid damage during penetration. It was inserted 30 cm
farther into the ground at the test depth. The shear test was run
immediately after insertion of the vane with the rotation speed
of about 61/min to obtain the undisturbed shear strength. The
remoulded shear strength was next measured after 15 rotations
of the vane, as previously performed by Horpibulsuk et al.
(2007). Thus, the sensitivity was obtained from the ratio of the
undisturbed vane shear strength to the remoulded vane shear
strength. Marine site investigation works were planned with
the aid of a geophysical seismic reflection survey. Boreholes
were drilled at locations with thick marine clay and other
locations to determine the marine clay profile. Due to the large
extent of the project area and variations in the underlying soil
profile, marine boreholes were created with offshore jack-up
pontoons. Fig. 2 presents the geological profile along Y 1000,
which is along the centre line of the runway, and Y 800, which
is along the centre line of taxiway 1 in the study area based on
the seismic profiling and the marine borehole results. Con-
tinuous undisturbed sampling holes were made at intervals of
every metre of depth, and continuous in-situ field vane tests
were carried out at intervals of 1 m of depth approximately 1 m
away from the locations of the sampling boreholes. The
termination criteria were based on achieving three consecutive
Standard Penetration Tests blow counts of 50.
3.2. Laboratory testing
Laboratory tests were performed on the recovered borehole
samples to determine the geotechnical properties and miner-
alogy of the marine clay. Oedometer consolidation tests were
conducted to determine the compression properties of the
Singapore marine clay at Changi. The oedometer specimens
were 63.5 mm in diameter and 19 mm in height. The con-
solidation load was applied in 24-h loading stages with a load
increment ratio of unity. Rowe cell tests (Rowe and Barden
(1966)) were undertaken to determine the coefficient of
consolidation due to the horizontal flow of the marine clay.
The employed Rowe cell was 75 mm in diameter and 30 mm
in thickness. The consolidation load was applied in 24-h stages
with a load increment ratio of unity.
The mineralogy of the marine clay was determined by the
extraction of piston samples for the X-ray diffraction (XRD)
and scanning electron microscope (SEM) analysis. Photo-
graphic identification of the marine clay was undertaken with
a high resolution camera.
3.2.1. X-ray diffraction
A representative portion of each sample was removed and
dispersed in distilled water using a reciprocal shaker by
treatment with ultrasound. The suspension was then sieved
on 63 μm, and the material that had passed through the sieve
was then placed in a measuring cylinder and allowed to stand.
In order to prevent flocculation of the clay crystals, 2 ml of
0.1 M ‘Calgon’ (sodium hexametaphosphate) was added to
each suspension. After a period dictated by Stoke's Law, a
nominal o 2 μm was then re-suspended in a minimum of
distilled water and pipette onto a ceramic tile in a vacuum
apparatus to produce an oriented mount.
The XRD analysis was conducted using a Philips PW1700
series diffractometer equipped with a cobalt-target tube and
operated at 45 kV and 40 mA. The clay mineralogy of the air-
dried and glycol-solvated samples was determined using
oriented mounts from 1.5–321 2θ with 0.481 2θ/min where θ
 M.W. Bo et al. / Soils and Foundations 55 (2015) 600–613
Fig. 2. Typical geological profile of study area.
was the diffraction angle. The diffraction data was analysed
using Philips APD1700 software coupled with a JCPDS
database running on a DEC MicroVax 2000 micro-computer
system.
3.2.2. Scanning electron microscope
Sub-specimens, approximately 2 cm  1 cm, were freeze
dried prior to the specimen preparation. This involved rapidly
freezing each sub-specimen in liquid nitrogen and then drying
it in an Edwards Modulyo Freeze Drier for approximately
24 h. Once dry, the specimens were carefully fractured to
produce a freshly exposed surface for the SEM examination.
These specimens were then mounted on aluminium stubs and
coated in a layer of carbon, approximately 25 nm thick, in a
carbon evaporation coater.
The SEM specimens were examined in an Energy Dispersive
X-ray Analysis (EDXA) system, which provided qualitative
chemical information from areas of interest. An accelerating
603
voltage of 20 kV was used throughout this study. The qualitative
chemistry of each clay mineral is identified on the list of
elements detected. The elements listed in brackets are listed in
the approximate order of decreasing concentration estimated
from the peak heights.
3.2.3. Photographic identification
Photographic identification of the marine clay obtained from
the marine sampling boreholes was carried out with a high
resolution camera. The marine clay specimens were extruded
and the entire length of the specimen was carefully cut and
subsequently opened by two different approaches, by cutting it
with a wire saw and by splitting it open without cutting it.
Opening the specimen, as such, exposed the inner core of the
specimen to identify the textures and laminations of the marine
clay. The colour of the marine clay was compared with colour
charts to describe the colour of the various layers of marine
clay. Photographic identification enables the featuring of the
604 M.W. Bo et al. / Soils and Foundations 55 (2015) 600–613
Table 1
Engineering properties of Singapore marine clay.
Parameters Upper marine clay
Βulk density (kN/m3) 14.2–15.7
Water content (%) 70–88
Liquid limit (%) 80–95
Plastic limit (%) 20–28
Initial void ratio 1.8–2.2
Specific gravity 2.6–2.72
Compression index 0.6–1.5
Recompression index 0.09–0.16
Coefficient of consolidation – vertical (m2/year) 0.47–0.6
Coefficient of consolidation – horizontal (m2/year) 2–3
Overconsolidation ratio 1.5–2.5
colours, textures and laminations of the marine clay. The
identification depths of the sand seams, organic material and
layers of past exposure to oxidation are also evident.
4. Test results
The range in physical and compression characteristics of the
upper, lower and intermediate marine clays obtained from the
laboratory tests is tabulated in Table 1. The results of the soil
investigation indicate the presence of two distinct layers,
namely, the upper marine clay layer and the lower marine
clay layer. The intermediate stiff clay layer is, in reality, the
desiccated layer of the lower marine clay. This desiccated layer
separates these two distinct marine clay layers. The soft upper
marine clay layer lies in a range of depth of 20–30 m below the
seabed. The intermediate stiff clay has a range in thickness of
3–5 m, and this predominantly stiff sandy silt or sandy clay
layer is sandwiched between the upper marine clay and the
lower marine clay. The lower marine clay lies in a range of
depth of 30–50 m below the seabed. The upper marine clay is
soft with undrained shear strength values ranging between 10
and 30 kPa. The undrained shear strength of the lower marine
clay is between 30 and 60 kPa. The sensitivity of the marine
clay at Changi varies from 3 to 8, which can be described as
highly sensitive marine clay.
The upper marine clay has a liquid limit of 80–95%, a
plastic limit of 20–28% and a water content of 70–88%. The
lower marine clay has a liquid limit of 65–90%, a plastic limit
of 20–30% and a water content of 40–60%. The intermediate
stiff clay has a liquid limit of about 50%, a plastic limit of
18–20% and a water content of 10–35%. The upper marine
clay is overconsolidated with a compression index of 0.6–1.5,
a coefficient of consolidation due to the vertical flow (cv) of
0.47–0.6 m2/year and a coefficient of consolidation due to the
horizontal flow (ch) of 2–3 m2/year. On the other hand, the
lower marine clay is slightly overconsolidated with a compres-
sion index of 0.6–1.0, a cv of 0.8–1.5 m2/year and a ch of
3–5 m2/year. The intermediate stiff clay is moderately over-
consolidated due to desiccation, with a compression index of
0.2–0.3, a cv of 1–4.5 m2/year and a ch of 5–10 m2/year. The
ch/cv ratio of the soft clay is very useful for determining the
Intermediate stiff clay Lower marine clay
18.6–19.6 15.7–16.6
10–35 40–60
50 65–90
18–20 20–30
0.7–0.9 1.1–1.5
2.68–2.76 2.7–2.75
0.2–0.3 0.6–1.0
0.08–0.15 0.14–0.2
1–4.5 0.8–1.5
5–10 3–5
3–4 2
rate of consolidation for preloading with the vertical drain
applications. It is found that the ch/cv values for the upper and
lower marine clay layers vary from 3.5 to 5.0. The upper
marine clay possesses a higher range of ch/cv values. The ch/cv
values vary from 4.25 to 5.0 for upper marine clay and 3.3 to
3.5 for lower marine clay. Bergado et al. (1992) reported that
the ch/cv ratio varies from 4 to 10 for Bangkok clay in
Thailand.
The SEM photomicrographs of the upper and lower marine
clays are shown in Fig. 3 (plates 1–6). Plates 1–3 are for the
upper marine clay at a depth of 6 m. Plates 4–6 are for the
lower marine clay at a depth of 15 m. Plates 1 and 4 show the
typical clay morphology. Plate 2 shows the typical clay
morphology and an open, poorly compacted texture. Plate 3
indicates the presence of rare framboical pyrite. Plate 5
indicates the presence of kaolinite particles within the clay
matrix. Plate 6 shows typical open, porous clay fabric. The
fabric of the marine clay at this depth of 15 m appears to be
generally open with porosities optically estimated at up to
30%. The calculated porosity of the clay with a water content
of 50% at this depth is 56%, which is greater than that
determined from the optical estimation. The lower optical
estimation might be due to the shrinkage of the specimen
during the freeze drying process. The clay specimens are
generally composed of ragged clay mineral flakes typically less
than 10 μm in diameter. The platelets repel each other weakly
when placed face to face. The connection between platelets
involves the edge of one platelet touching the centre of the face
of another. The resulting configuration resembles a house of
cards as well as poorly developed clay flakes. Rare detrital
muscovite flakes, quartz and K-feldspar silt grains and shell
fragments are distributed throughout the clay matrix. Authi-
genic pyrite occurs as isolated sub-spherical particles and rare
framboids, suggesting an anoxic diagenetic environment.
Poorly developed clay flakes, apparently associated with the
pyrite, have a Fe, K, Mg, Al, Si chemistry and may be
composed of chlorite. These chemical components cause
cementation bonding between the clay particles. EDXA
indicates that the clay has K, Fe aluminium silicate, dominat-
ing in most specimens. Such a chemical composition probably
corresponds to the kaolinite mineral. Rarer flakes with Mg, K,
 M.W. Bo et al. / Soils and Foundations 55 (2015) 600–613 605

Fig. 3. SEM photomicrographs of marine clay: (a) saw-cut open samples and (b) broken-open sample.

Fe (tr Ca) aluminium silicate to K, Fe (tr Ca, Ti) aluminium
silicate composition may relate to the smectite-group mineral.
Figs. 4–6 present the textures of the clay specimens. The
left-hand side photos show the texture of the clay specimen,
which is cut open with a wire saw where some smearing has
occurred. The right-hand side photos show the texture of the
clay specimen when it is spilt open without cutting. Fig. 4
shows the photographic identification of the upper marine clay
which is brownish-blue with organic deposits. Fig. 5 shows the
photographic identification of the intermediate stiff clay which
is reddish due to the oxidation of the layer. This indicates the
results of the exposure of the seabed to the atmosphere during
the rise and fall of the sea levels in the geological past. Fig. 6
shows the photographic identification of the lower marine clay
which is brownish-blue with organic deposits. Fine sand
particles are observed in all the layers. The XRD results for
the specimens at depths between 2 and 18 m are shown in
Fig. 7. The strong reflection at 14.412θ for all samples
indicates the presence of kaolinite as the main clay mineral;
this supports the SEM and EDXA results. The peaks between
6.31 2θ and 6.91 2θ represent smectite, and the peak at 101 2θ
represents mica.
606 M.W. Bo et al. / Soils and Foundations 55 (2015) 600–613

Fig. 4. Photographic identification of upper marine clay: plates 1–3 for upper marine clay at depth of 6 m and plates 4–6 for lower marine at depth of 15 m.
(For interpretation of the references to colour in this figure the reader is referred to the web version of this article.)

Fig. 5. Photographic identification of intermediate stiff clay. (For interpretation of the references to colour in this figure, the reader is referred to the web version of
this article.)

5. Analysis of mineralogy and geotechnical properties
The typical soil profile and geotechnical properties (includ-
ing water content, physical properties, vane shear strength,
sensitivity and compression characteristics) of a boring log at
Changi are shown in Fig. 8. The liquid limits are within the
range of 40–90% and predominantly greater than 50%. The
water contents are very close to the corresponding liquid limits
(liquidity index close to 1.0) at depths of 0–10 m. For deeper
depths, the water contents are slightly lower than the corre-
sponding liquid limits. The specific gravities are almost similar
for the entire depth, being approximately 2.6 7 0.1. The
sensitivities vary from 2 to 7 with an average value of 5.
The field vane shear strength increases linearly with depth at
depths of o 15 m. The field vane shear strength varies from
35 kPa to 65 kPa for deeper depths. The activity reduces with
depths up to 8 m and then increases with depth. The activity is
in the range of 0.5–1.3; thus, the Singapore marine clay is
considered as inactive to active, but is mostly inactive.
As is evident from Fig. 9 that the liquid limit (LL), plasticity
index (PI) and activity vary significantly with depth, and that
the PI and LL points from different depths are above the A-line
 M.W. Bo et al. / Soils and Foundations 55 (2015) 600–613
Fig. 6. Photographic identification of lower marine clay. (For interpretation of the references to colour in this figure, the reader is referred to the web version of this
article.)
on the plasticity chart. It is observed that the PI of Singapore
marine clay bears a functional relationship to LL. Fig. 10
shows the relationship between PI and the clay fraction of
Singapore clay with various depths compared with that of
Bangkok clay reported by Horpibulsuk et al. (2007). The test
data on Bangkok clay was selected for comparison with that of
Singapore clay because both Singapore and Bangkok clays are
Southeast Asian marine clays and are classified as problematic
clays. The slope of the relationship represents the soil activity,
which is mainly controlled by the clay mineralogy and pore
fluid (Skempton, 1953). Fig. 10 suggests that the activity of
Bangkok clay is higher than that of Singapore clay. The
majority of activity values for Bangkok clay are greater than
1.25, indicting it as active clay, while the majority of activity
values for Singapore clay are lower than 1.25.
The clay minerals of Singapore marine clay are summarised in
Fig. 11 and compared with those of Bangkok clay (Ohtsubo et al.
(2000)) to understand the role of clay mineralogy in the index and
geotechnical properties of remoulded Singapore clay. The quantity
of each clay mineral was determined from the ratio of the area
under the XRD curve for the considered mineral (kaolinite,
smectite or mica) to the total area under the XRD curve for all
clay minerals. For Singapore clay, the kaolinite content is in a
range of 62–70%, followed by smectite and mica for the whole
depth. On the other hand, Bangkok clay contains high smectite
content, followed by kaolinite and mica. The kaolinite content in
Singapore marine clay is approximately twice higher than that of
Bangkok clay for all depths. Since kaolinite is a non- to low-
swelling mineral, the activity and liquid limit of Singapore clay are
found to be significantly lower than those of Bangkok clay. The
liquid limit of Singapore clay is between 30% and 70%, while it
was reported to be between 70% and 110% for Bangkok clay at
Bangpee (Horpibulsuk et al., 2007). The effect of the liquid limit
on the compression behaviour is shown in Section 6.
607
6. In-situ stress state of Singapore clay
The in-situ stress state of clay can be examined using the
intrinsic (reference) state line (Nagaraj et al., 1998; Nagaraj
and Miura, 2001; Horpibulsuk et al., 2007, 2011). The
Intrinsic State Line (ISL) for a clay–water system devoid of
any stress history, time or cementation effects is the path
obtained by the compression of remoulded clay. The compres-
sion behaviour of remoulded Singapore clay samples with
different liquid limits is illustrated in Fig. 12. The difference in
the compression index is the result of the variation in fabric
patterns, which control the response of the clay (Nagaraj et al.,
1990). The specimen with a higher liquid limit exhibits a
higher slope (compression index). Since clays at the liquid
limit have the same order of pore-water suction (5–6 kPa)
(Russell and Mickle, 1970; Wroth and Wood, 1978),
undrained shear strength (1.7–2.5 kPa) and hydraulic conduc-
tivity (10
7 cm/s) (Nagaraj et al., 1993), the liquid limit has
been successfully used as a state parameter to obtain the ISL
(Nagaraj et al., 1998). The ISL was empirically proposed based
on 11 different clays with various liquid limits ranging from
36% to 159%. As such, the validity of this proposed ISL needs
to be examined for the clay studied here. The ISL for
Singapore marine clay is presented, as shown in Eq. (1), and
compared with that proposed by Nagaraj et al. (1998) in
Fig. 12
eR 0
¼ 1:226
0:286 log σ v ð1Þ
eL
where eR is the void ratio in the remoulded state and eL is the
void ratio at the liquid limit. This relation has the same pattern
as that proposed by Nagaraj et al. (1998); Horpibulsuk et al.
(2007), and yet the slope for Bangkok clay is found to be
slightly higher than that of remoulded Singapore clay.
608 M.W. Bo et al. / Soils and Foundations 55 (2015) 600–613

The ISL is useful for analyses of the natural stress state and
the compression behaviour of natural soils whose formation
depends on various physical, chemical and environmental
properties (Burland, 1990; Horpibulsuk et al., 2007, 2011).
The in-situ state of Singapore clay (upper, intermediate and
lower marine clay layers), with reference to the ISL (Eq. (1)),
0
is presented in Fig. 13. It is shown that the e0 and σ v0 points of
Singapore clay are located above the ISL for all three layers,
0
where e0 and σ v0 are the in-situ void ratio and in-situ effective
vertical stress, respectively. This finding indicates that natural
Singapore marine clay is stable in the meta-stable state and
designated as a “structured clay” (Burland, 1990; Mitchell,
1996). This is because the initial fabric of natural Singapore
clay is open and involves some amount of edge-to-edge and
edge-to-face associations in a “cardhouse” arrangement and the
particle contacts are connected by the chemical bonds (as
presented earlier in Fig. 3).

7. Modelling of compression curve of Singapore clay

Based on the analysis of the compression curves of various

Fig. 7. XRD patterns of air dried and glycol-solvated samples for different
depths.
undisturbed clays, the idealised compression behaviour of
natural structured clay has been proposed, as presented in
Fig. 14 (Liu and Carter, 1999). The compression curve of
natural structured clay lies above the destructured (remoulded)
line even at very high effective vertical stress. The compres-
0
sion strain is small up to the yield stress, σ y , due to the
contribution of the clay structure. Beyond this yield stress,
there is sudden compression of a relatively high magnitude,
indicated by the steep slope and caused by the destructuring.
On further loading, the difference in void ratio between the
structured and the remoulded clays (es), with logarithm of
stress, decreases to an almost constant value. This constant
value is designated as the residual additional void ratio esr. It is
evident from this study that the compression curve of natural
Singapore clay is consistent with the idealised compressive
curve in Fig. 14.
From Fig. 14, the void ratio at a particular effective vertical
stress can be expressed (Liu and Carter, 1999) as
e ¼ e R þ es ð2Þ
where e represents the void ratio for structured clay and es is
the additional void ratio. The soil structure is defined as fabric
(arrangement of clay particles and pore space) and interparticle
forces (Nagaraj and Miura, 2001; Mitchell, 1996). The eR is
the void ratio supported by the intrinsic soil structure and
dependent upon the clay mineral and the pore fluids, while es is
the additional void ratio attributed to the ageing effect (stress
history and cementation).
Liu and Carter (1999, 2000), Horpibulsuk et al. (2007, 2010,
2013) and Suebsuk et al. (2010, 2011) have analysed the
compression behaviour of natural and artificially structured
clays and have found that the additional void ratio due to the
ageing effect, es, at the post-yield state is proportional to
0
σ y =ðσ v Þb , where σy is the yield stress and b is the destructuring
index, quantifying the rate of destructuring. The greater ageing
 M.W. Bo et al. / Soils and Foundations 55 (2015) 600–613
Fig. 8. Soil profile and geotechnical properties of Singapore marine clay.
Fig. 9. Plasticity index versus liquid limit of Singapore marine clay.
Fig. 10. Activity of Singapore clay and Bangkok clay.
609
effect results in the higher σy. The greater ageing effect is also
associated with higher values for b due to the sudden
destructuring (Horpibulsuk et al., 2004, 2010; Du et al.,
2014). The additional void ratio is expressed (Liu and
Carter, 1999) as follows:
0
!b
σy
es ¼ esy 0 þ esr ð3Þ
σv
where esy is the additional void ratio corresponding to the yield
stress and esr is the residual additional void ratio that cannot be
eliminated by an increase in the effective vertical stress (when
0
σ v approaches infinity, es ¼ esr). The b value can be determined
by the rate of the change in void ratio over the change in
effective vertical stress by considering the power relationship
described in Eq. (3). The b value can also be determined by
fitting because the values for all the other parameters can be
determined based on their physical meanings. Based on the test
results for Singapore marine clay, b is equal to 1.0, which is
similar to that for Bangkok clay. The esy can be determined by
0 0
considering that σ v ¼ σ y , as shown in the following equation:
esy þ esr ¼ ey
eRy ð4Þ
where ey and eRy are the void ratios of the structured and
remoulded curves corresponding to the yield stress, respec-
tively. Based on this destructuring framework, the experimen-
tal virgin compression curve of Singapore clay can be
0
modelled once σ y and ey are known. Following is the stepwise
procedure for determining the compression curve of the
Singapore marine clay:
1. Determine the in-situ void ratio, overburden stress, index
properties, compression curve and yield stress of the
undisturbed sample.
0
2. Draw a remoulded (e
log σ v ) line, which can be directly
obtained from consolidation tests on remoulded clay or
estimated from Eq. (1).
3. Determine the esr and compression index at the pre-yield
state, Cs.
610 M.W. Bo et al. / Soils and Foundations 55 (2015) 600–613

Fig. 11. Clay mineral composition and activity of Bangkok clay and Singapore clay.

Fig. 14. Compression behaviour of structured clays (modified from Nagaraj

et al. (1990) and Liu and Carter (1999)).

Fig. 12. Compression behaviour for different liquid limits and intrinsic
state lines.
4. From the known Cs, determine the void ratio at the yield
stress and draw a compression path at the pre-yield state.
5. Draw a compression path at the post-yield state using
Eqs. (2)–(4).

Fig. 15 shows an example of the prediction of the experi-

Fig. 13. Natural stress state of Singapore clay with reference to intrinsic
state line.
mental compression curves of Singapore clays for the three
layers. The remoulded lines were approximated using Eq. (1).
The predicted and measured compression curves show a
reasonably good agreement. From the results shown in
Fig. 15 and an analysis of the extensive data, it is revealed
that esr is practically constant, at about 0.11 for Singapore clay
with three layers, while esr for soft Bangkok clay is 0.20. The
difference in esr is possibly due to the different soil structures.
The lower esr of the Singapore clay implies that the ageing
effect of Singapore clay is less than that of Bangkok clay.
Suebsuk et al. (2010) and Horpibulsuk and Liu (2015)
illustrated that stronger cementation bonds (interparticle force)
 M.W. Bo et al. / Soils and Foundations 55 (2015) 600–613 611

Fig. 15. Predicted and measured compression curves of Singapore marine clay.
612 M.W. Bo et al. / Soils and Foundations 55 (2015) 600–613
result in higher esr, i.e., esr of cement-stabilized clay is
significantly higher than that of natural clay.
8. Conclusions
The Singapore marine clay used in this engineering geology
study, undertaken in the Eastern part of the Republic of
Singapore, comprises a soft upper marine clay layer overlying
a stiffer lower marine clay layer. An intermediate stiff clay
layer is sandwiched between these two marine clay layers,
which was the desiccated crust of the lower marine clay layer.
The depth of the soft upper marine clay layer ranges from 20 to
30 m below the seabed. The thickness of the intermediate stiff
clay lies in the range of 3–5 m and this predominantly stiff
sandy silt or sandy clay layer is sandwiched between the upper
marine clay and lower marine clay. The depth of the lower
marine clay ranges from 30 to 50 m below the seabed.
The upper marine clay is soft, with undrained shear strength
values ranging from 10 to 30 kPa. The undrained shear
strength of the lower marine clay is between 30 and 60 kPa.
The sensitivity of the marine clay at Changi varies from 3 to 8,
which is described as highly sensitive marine clay. The upper
clay layer has a cv of 0.47–0.6 m2/year and a ch of 2–3 m2/
year. The lower marine clay has a cv of 0.8–1.5 m2/year and a
ch of 3–5 m2/year, and the intermediate stiff clay has a cv of
1–4.5 m2/year and a ch of 5–10 m2/year.
The fabric of the marine clay appears to be generally open
with “cardhouse” characteristics. The XRD analysis indicates
the major content of minerals to be kaolinite and smectite with
“mica” and chlorite being the minor minerals. Photographic
identification of Singapore marine clay indicates the brownish-
blue upper and lower marine clay layers, consisting of organic
deposits and fine sand particles. The intermediate stiff clay
layer is reddish due to oxidation of the layer as a result of the
exposure of the seabed to the atmosphere during the rise and
fall of the sea levels in the geological past.
Compared with Bangkok clay, the liquid limit and the
activity of the Singapore clay are lower because of the higher
kaolinite content. The marine Singapore clay is classified as
inactive to active clay, but is mostly inactive clay. The
compression curves of the remoulded Singapore clay are
different for different liquid limits. Using the liquid limit as
the state parameter, the Intrinsic State Line (ISL) for Singapore
marine clay is developed. With open fabric and chemical
bonds, the natural states of Singapore marine clay for the three
layers are in the meta-stable state and are designated as
structured clay. The compression curves of the natural
Singapore clay for various depths can be modelled using the
ISL and the destructuring framework. This modelling will be
useful for settlement calculations and the development of a
future constitutive model for Singapore marine clay.
Acknowledgements
This research was supported under the funding theme of the
Australian Research Council's Linkage Projects (Project num-
ber LP0989164). The third and fourth authors are grateful for
the financial support received from the Thailand Research
Fund, under the TRF Senior Research Scholar programme,
Grant no. RTA5680002.
References
Abuel-Naga, H.M., Bergado, D.T., Bouazza, A., Pender, M.J., 2009. Thermal
conductivity of soft Bangkok clay from laboratory and field measurements.
Eng. Geol. 105, 211–219.
Arulrajah, A., Nikraz, H., Bo, M.W., 2004a. Observational methods of
assessing improvement of marine clay. Proc. Inst. Civil Eng. – Ground
Improv. 8 (4), 151–169.
Arulrajah, A., Bo, M.W., Nikraz, H., Hashim, R., 2004b. Piezocone dissipation
testing of singapore marine clay at Changi. J. Southeast Asian Geotech.
Soc. – Geotech. Eng. 35 (3), 119–126.
Arulrajah, A., Nikraz, H., Bo, M.W., 2005. In-Situ testing of singapore marine
clay at Changi. Geotech. Geolog. Eng. 23, 111–130.
Arulrajah, A., Nikraz, H., Bo, M.W., 2006a. Assessment of marine clay
improvement under reclamation fills by in-situ testing methods. Geotech.
Geolog. Eng. 24 (1), 219–226.
Arulrajah, A., Nikraz, H., Bo, M.W., 2006b. In-situ pore water pressure
dissipation testing of marine clay under reclamation fills. Geotech. Geolog.
Eng. 24 (1), 29–43.
Arulrajah, A., Bo, M.W., Nikraz, H., Balasubramaniam, A.S., 2007. Dissipa-
tion testing of singapore marine clay by piezocone tests. J. Geotech.
Geolog. Eng. 25 (1), 647–656.
Arulrajah, A., Abdullah, A., Bo, M.W., Bouazza, A., 2009a. Ground
improvement techniques for railway embankments. Proc. Inst. Civil
Engineers – Ground Improv. 162 (GI1), 3–14.
Arulrajah, A., Bo, M.W., Chu, J., Nikraz, H., 2009b. Instrumentation at the
Changi Land Reclamation Project, Singapore. Proc. Inst. Civil Eng. –
Geotech. Eng. 162 (1), 33–40.
Arulrajah, A., Bo, M.W., Piratheepan, J., Disfani, M., 2011. In-situ testing of
soft soil at a case study site with the self-boring pressuremeter. ASTM
Geotech. Test. J. 34 (4), 355–363.
Baudet, B.A., Stallebrass, S.E., 2001. Modelling the destructuring of soft
natural clays. Comput. Methods Adv. Geomech. 1, 297–302.
Bergado, D.T., Enriquez, A.S., Sampaco, C.L., Alfaro, M.C., Balasubrama-
niam, A.S., 1992. Inverse analysis of geotechnical parameters on improved
soft Bangkok clay. J. Geotech. Eng. Div. 118 (7), 1012–1030.
Bo, M.W., Chu, J., Low, B.K., Choa, V., 2003. Soil Improvement –
Prefabricated Vertical Drain Techniques. Thomson Learning, Singapore,
068808.
Bo, M.W., Chang, M.F., Arulrajah, A., Choa, V., 2000. Undrained shear
strength of the Singapore marine clay at Changi from in-situ tests. Geotech.
Eng. J. Southeast Asian Geotech. Soc. 31 (2), 91–107.
Bo, M.W., Arulrajah, A., Leong, M., Horpibulsuk, S., Disfani, M.M., 2013.
Evaluating the in-situ hydraulic conductivity of soft soil under land
reclamation fills with the BAT permeameter. Engineering Geology 168,
98–103.
Burland, J.B., 1990. On the compressibility and shear strength of natural clays.
Geotechnique 40 (3), 329–378.
Chai, J.C., Miura, N., Zhu, H.H., 2004. Compression and consolidation
characteristics of structured natural clays. Can. Geotech. J. 41 (6),
1250–1258.
Chu, J., Bo, M.W, Chang, M.F., Choa, V., 2002. Consolidation and
permeability properties of Singapore marine clay. J. Geotech. Geoenviron.
Eng. 128 (9), 724–732.
Du, Y.J., Horpibulsuk, S., Wei, M.L., Suksiripattanapong, C., Liu, M.D., 2014.
Modeling compression behavior of cement treated zinc contaminated
clayey soils. Soils Found. 54 (5), 1018–1026.
Hong, Z., Onitsuka, K., 1998. A method of correcting yield stress and
compression index of Ariake clays for sample disturbance. Soils Found. 38
(2), 211–222.
Horpibulsuk, S., Liu, M.D., 2015. Structured Cam Clay model with cementa-
tion effect. Geotechnical Engineering Journal of the SEAGS & AGSSEA
46 (1), 86–94.
 M.W. Bo et al. / Soils and Foundations 55 (2015) 600–613
Horpibulsuk, S., Bergado, D.T., Lorenzo, G.A., 2004. Compressibility of
cement-admixed clays at high water content. Geotechnique 54 (2),
151–154.
Horpibulsuk, S., Rachan, R., Suddeepong, A., Liu, M.D., Du, Y.J., 2013.
Compressibility of lightweight cemented clays. Eng. Geol. 159, 59–66.
Horpibulsuk, S., Yangsukaseam, N., Chinkulkijniwat, A., Du, Y.J., 2011.
Compressibility and permeability of Bangkok clay compared with kaolinite
and bentonite. Appl. Clay Sci. 52, 150–159.
Horpibulsuk, S., Liu, M.D., Liyanapathirana, D.S., Suebsuk, J., 2010.
Behaviour of cemented clay simulated via the theoretical framework of
the structured cam clay model. Comput. Geotech. 37 (1–2), 1–9.
Horpibulsuk, S., Shibuya, S., Fuenkajorn, K., Katkan, W., 2007. Assessment of
engineering properties of Bangkok clay. Can. Geotech. J. 44 (2), 173–187.
Jung, Y-H., Finno, R.J., Cho, W., 2012. Stress–strain responses of
reconstituted and natural compressible Chicago glacial clay. Eng. Geol.
129–130, 9–19.
Karakouzian, M., Avar, B.B., Hudyma, N., Moss, J.A., 2003. Field measure-
ments of shear strength of an underconsolidated marine clay. Eng. Geol.
67, 233–242.
Kim, D., Kim, K-S., Ko, S., Choi, Y., Lee, W., 2012. Assessment of
geotechnical variability of Songdo silty clay. Eng. Geol. 133–143, 1–8.
Liu, M.D., Carter, J.P., 2000. Modelling the destructuring of soils during virgin
compression. Géotechnique 50 (4), 479–483.
Liu, M.D., Carter, J.P., 1999. Virgin compression of structured soils.
Geotechnique 49 (1), 43–57.
Liu, M.D., Xu, K.J., Horpibulsuk, S., 2013. A mathematic function for the s-
shape relationship for geotechnical applications. Proc. ICE – Geotech. Eng.
166 (3), 321–327.
Liu, M.D., Carter, J.P., Airey, D.W., 2011. Sydney Soil model: (I) theoretical
formulation. Int. J. Geomech. 11 (3), 211–224.
Mitchell, J.K., 1996. Fundamentals of Soil Behaviour. John Willey & Sons
Inc., New York, 437.
Nagaraj, T.S., Miura, N., 2001. Soft Clay Behaviour: Analysis and Assess-
ment. Balkema, Rotterdam, 315.
613
Nagaraj, T.S., Pandian, N.S., Narasimha Raju, P.S.R., 1998. Compressibility
behavior of soft cemented soils. Géotechnique 48 (2), 281–287.
Nagaraj, T.S., Pandian, N.S., Narasimha Raju, P.S.R., 1993. Stress state –
permeability relationships for fine-grained soils. Géotechnique 43 (2),
333–336.
Nagaraj, T.S., Srinivasa Murthy, B.R., Vatsala, A., Joshi, R.C., 1990. Analysis
of compressibility of sensitive clays. J. Geotech. Eng. 116 (GT1), 105–118.
Ohtsubo, M., Egashira, K., Koumoto, T., Bergado, D.T., 2000. Mineralogy and
chemistry, and their correlation with the geotechnical index properties of
Bangkok clay: comparison with Ariake clay. Soils Found. 40 (1), 11–21.
Pestana, J.M., Whittle, A.J., 1995. Compression model for cohesionless soils.
Géotechnique 45, 621–631.
Potts, D.M., Zdravkovic, L., 1999. Finite Element Analysis in Geotechnical
Engineering: Theory. Thomas Telford, London.
Public Works Department Singapore, 1976. The Geology of Singapore Public
Works Department Report. Public Works Department, Singapore.
Rowe, R.W., Barden, L., 1966. A new consolidation cell. Geotechnique 26 (2),
162–170.
Russell, E.R., Mickle, J.L., 1970. Liquid limit values of soil moisture tension.
J. Soil Mech. Found. Div. 96, 967–987.
Skempton, A.W., 1953. Activity of clays. In: Proceedings of the 3rd
International Conference on Soil Mechanics Foundation Engineering,
vol. 1, pp. 195–200.
Suebsuk, J., Horpibulsuk, S., Liu, M.D., 2010. Modified structured cam clay: a
constitutive model for destructured, naturally structured and artificially
structured clays. Comput. Geotech. 37 (7–8), 956–968.
Suebsuk, J., Horpibulsuk, S., Liu, M.D., 2011. A critical state model for
overconsolidated structured clays. Comput. Geotech. 38 (5), 648–658.
Wroth, C.P., Wood, D.W., 1978. The correlation of index properties with some
basic engineering properties of soils. Can. Geotech. J. 15 (2), 137–145.
Yan, W.M., Ma, Y., 2010. Geotechnical characterization of Macau marine
deposits. Eng. Geol. 113, 62–69.