CHAPTER ONE

INTRODUCTION

1.0) Background

During the past two decades there have been many major civil engineering

projects in Singapore. The most important of these have been high rise buildings, rec-

lamation projects, MRT tunnels, expressways, and expansion of the airport and port

facilities. This rapid rate of construction to cater for the fast growing industrial and

commercial sectors together with residential developments in Singapore have led to

ever increasing complexities in design and the resulting need for a more accurate de-

scription of soil behaviour. Accurate determination of appropriate soil parameters is

important in describing the in-situ soil behaviour. However, the current understanding

of the geotechnical behaviour and properties of Singapore’s soil deposits is still rather

rudimentary (Chang, 1991; Pitts, 1984a; Yong et al., 1990). To the author’s awareness

there has not been any published detailed soil investigation that has being carried out

on the in-situ properties of Singapore soil in the last 15 years. Most of the geotechni-

cal project’s design parameters used by local geotechnical engineers are correlated to

basic indices such as Atterberg Limits and undrained shear strength, which are ob-

tained using conventional testing methods (Dames and Moore, 1983; Tan, 1983).

These parameters cannot describe in-situ soil behaviour to a degree that is compatible

with the advancement in understanding soil behaviour today and the considerably more

sophisticated finite element analyses often carried out for design.

The growing need to simulate actual construction processes more accurately

has led to a rise in the use of the finite element method (Chew et al., 1996; Lee et al.,

1997). Such analyses usually require the use of an appropriate constitutive model with

1
parameters that could be ascertained from appropriate soil investigations. As more and

more constructions are carried out in heavily built up areas, there is a growing need to

carry out deformation analyses accurately. To cater for such analyses, a chosen model

must be able to characterise the in-situ behaviour of the soil in terms of non-linearity

and recent stress history (Atkinson et al., 1990; Stallebrass, 1990 Viggiani, 1992; Vig-

ianni and Atkinson, 1995; Stallebrass et al., 1997; Dasari, 1996). In-situ soil parame-

ters for the chosen constitutive law are often determined using some field or laboratory

test data. The parameters must be able to simulate reasonably the in-situ soil behav-

iour, which can only be determined from a systematic study.

A further problem, particularly severe in soil mechanics, is the global and re-

gional variation in soil types and behaviour. Marine clay in Singapore can display a

very different behaviour from that in Japan or Europe due to the different depositional

and environmental histories. Extensive studies have been conducted on some soils

such as Ariake clay in Japan (Ohtsubo et al., 1985; Hanzawa et al., 1990; Mukabi,

1995; Mitachi et al., 1996; Tanaka et al., 1996a; Tanaka and Tanaka, 1997), Bothken-

nar clay in Scotland (Hawkin et al., 1989; Hight et al., 1992a; Atkinson et al., 1992;

Leroueil et al., 1992; Clayton et al., 1992; Nash et al., 1992b), Drammen clay in Nor-

way (Lunne et al., 1976; Lacasse and Lunne, 1982; Lacasse et al., 1985; Lunne, 1997)

and Champlain clay in Canada (Locat et al., 1984; Morin et al., 1983; Lefebvre et al.,

1991; Lefebvre et al., 1994; Leroueil et al., 1995). To date, no such study have been

conducted on geological deposit covering the Singapore Island. In recent years, con-

siderable efforts have been devoted to numerical analysis method (Chew et al., 1996;

Wang et al., 1997; Lee et al., 1997; Lee et al., 1998) and physical modelling using

centrifuge (Lee et al., 1991; Leong et al., 1994; Goh, 1994; Ng et al., 1994; Yet et al.,

1994; Niu et al., 1997). Results from numerical modelling are dependent on the input

2
soil properties and soil model used, especially for the in-situ situation. Physical mod-

elling, while useful, often cannot reproduce the full spectrum of the in-situ soil behav-

iour due to the fact that remoulded soil is used. In order to allow numerical modelling

to achieve its full potential, it is important that a proper characterisation study is car-

ried out incorporating many of the recent advances in geotechnical testing that have yet

to be used on a Singapore soil. In addition, the author is also aware of the importance

of the effect of sample quality, an aspect that has not received much attention in Singa-

pore. Besides the general information given above, there is few published literature on

issues such as small strain behaviour, compressibility behaviour, yielding characteris-

tic, detailed stress strain behaviour and the effect of sampling disturbance on all these

parameters.

1.1) Introduction of The Geological History of Singapore Marine Clay

The geological deposits on Singapore Island is usually divided into six major

formations known locally as Kallang Formation, Old Alluvium, Jurong Formation,

Bukit Timah Granite, Gombak Norite and Sahajat Formation (Fig. 1.1) (Yong and

Karunaratne, 1983; Pitts, 1984a; Pitts, 1992) The Kallang Formation covers much of

the coastal plain, immediate offshore zone and the deeply incised river valleys, which

penetrate to the center of Singapore Island. The most important member of the Kal-

lang Formation is the marine clay. The Old Alluvium occurs mainly in the eastern part

of the Singapore Island where it exists as a virtually uninterrupted sheet either at the

surface or underneath younger deposits. The Jurong Formation comprises weathered

sedimentary rocks such as sandstones, siltstones and mudstones of Jurassic age, and

covers most of the western part of Singapore. The Bukit Timah Granite covers over

one-third of the surface of central Singapore. Finally, the Lower Palaezoic sandstones

3
and mudrocks of the Sahajat Formation crop out in one small area only. They are

known to exist beneath younger deposits in the eastern part of Singapore Island. So

far, no detailed soil investigation has been carried on any of these soils. The current

study can be viewed as an initial exercise to characterise Singapore soils.

The Kallang Formation is a recent deposit in Singapore, which consists of soil

of marine, alluvial, littoral, and estuarine origins (peaty soil). These deposits, which

are widely distributed both on the main Singapore island and offshore, cover nearly

25% of the total land surface of the Singapore island (Chang, 1991; Yong et al., 1990)

as shown in the geological map of Singapore in Fig. 1.1. This formation is named after

its extensive occurrence in the Kallang River Basin. Kallang Formation generally con-

sists of two clay members separated by a stiffer and sandy intermediate layer (Fig.

1.2). These two clays are usually called the Upper and Lower Marine Clay. In some

areas, organic or peaty soil is present at the base of the upper member, this being a

transition member of the Kallang Formation (Pitts, 1984b; Pitts, 1992). The formation

varies in thickness from place to place, usually between 10 m to 15 m near the estuar-

ies. But in some instances, it can be as thick as 40 m (Yong et al., 1983).

There are two schools of thoughts on the formation of this two clay formation

separated by a stiffer sandy intermediate layer. One school postulated that this inter-

mediate layer is the weathered and desiccated crust of the lower marine clay. One

strong basis for this argument is shown in Fig. 1.3 which shows the sea level changes

in the post-glacial period postulated by Fairbridge (1961) and Kenney (1964), and to-

gether with depositional history of Singapore marine clay presented by Pitts (1984a

and 1992). The lower marine member was deposited non-uniformly over valley and

plain floors some 12,000 years ago. This school argued that when the sea level

dropped by 20 m to 25 m during two regressions between 10,000 and 12,000 years

4
ago. As a result of the exposure of the lower member to desiccation and weathering, a

stiff weathered crust was formed on top of the lower member. The upper marine

member was subsequently deposited over this layer during a period that stretched from

10,000 years ago, when the sea level began to rise from the elevation of approximately

- 38 m after the two major regressions, to between 6,000 and 5,000 years ago, when the

sea reached the recorded maximum elevation of +2 m during the post-glacial Climatic

Optimum (Pitts, 1983 and 1992).

The second school believed that the intermediate layer is a fluvial deposit. Pa-

laeoenvironmental study by Poh et al. (1985) showed the presence of fresh water pteri-

dophyte spores suggesting a fresh water environment of deposition with proximity to

the brackish influence of mangrove swamp. The fluctuation in sea levels known to be

attributed to the freezing and thawing of glacial ice during the Quaternary Ice Age was

perhaps the cause of the change in depositional environment from marine to fresh wa-

ter. Therefore, the palaeoenvironmental study suggests that the intermediate layer of

stiff clay, existing between the upper and lower marine members, is a recent alluvial

deposit. This is also the view of Dames & Moore (1983) and others that it is a fluvial

deposit based on field observation in boreholes as opposed to the view that the inter-

mediate layer represents a weathered and desiccated crust of the Lower marine clay.

However, the precise origin of this layer is not critical to this study which focuses on

the Lower Marine Clay Member of the Kallang Formation.

According to one study by Tan (1983), Singapore marine deposit is essentially

a weakly flocculated, kaolinite- rich clay with moderate contents of montmorillonite

and illite. However, Lim (1982) noted that this mineralogical content could only be

confirmed on isolated samples. On the other hand, Cox (1968) believed that the domi-

nant clay mineral in the marine clay of Southeast Asia is illite. More than 50 recent

5
latest X-ray tests by Laval University, Canada indicates that the predominant mineral

is kaolinite with traces of illite and smectite (Tan et al., 1999; Tanaka et al., 2000; Tan

et al., 2003). These amounts of test are sufficient to verify the mineralogy of Singapore

marine clay. The clay content is generally 55% to 70% (Chong et al., 1998; Tan et al.,

1999; Tan et al., 2003) and the activity is around 0.95 for the upper and lower member

(Lim, 1982; Chong et al., 1998; Tan et al., 1999; Tan et al., 2003). Typically, the spe-

cific gravity is 2.50 to 2.75 for both members (Cheong and Subrahmanyam, 1979; Tan,

1983; Yong et al., 1990, Chong et al., 1998; Tan et al., 1999; Tan et al., 2003) and the

organic content is 3%, similar to the 3% to 5% reported by Cox (1968). The marine

member is normally pale grey to bluish grey in colour. It is fairly homogeneous with

occasional occurrences of shell fragments. The clay is highly plastic (typically LL=65

- 120, PI=20 - 45). The average bulk unit weight is 15.2 + 0.6 kN/m3, where the lower

marine clay has the higher bulk unit weight than upper marine clay. Its sensitivity

ranges approximately from 1.5 to 6 for upper marine clay and from 3.0 to 5.0 for lower

marine clay. With its natural moisture content at 50% to 90%, the corresponding li-

quidity index (LI) is approximately 0.60 to 1.0 (Tan, 1983; Yong et al., 1990).

The detailed variation of the OCR profile in clay deposit is directly related to

the post-depositional history of the clay. In the case of ageing, the clay deposit is usu-

ally characterised by constant OCR with depth (Bjerrum, 1967). Lowering of ground-

water or surface erosion would introduce a change in vertical effective stress. Conse-

quently, there would be a gradual decrease in OCR with depth. The effect of weath-

ering due to downward leaching or desiccation would normally lessen as the depth in-

creases and, as a result, the OCR would also decrease, perhaps at a fairly rapid rate,

with depth. The Singapore marine clay is generally considered to be slightly ”aged”

with OCR ranging from 1.1 to 1.6 according to Lim (1982). Fig. 1.4 shows a typical

6
OCR profile of the Singapore marine clay presented by Hanzawa and Adachi (1983).

Singapore marine clay has a compression index Cc of 0.7 to 1.3 for upper marine clay

and 0.45 to 0.95 for the lower marine clay (Tan, 1981; Tan, 1983; Yong et al., 1990),

whereas the recompression index ranges from 0.083Cc to 0.3Cc for both the upper and

lower marine clay (Tan, 1983; Yong et al., 1990). This ratio is considered low com-

pared to other clays such as Ariake clay and Bothkennar clay.

The above review indicates only soil indices, shear strength and compressibility

were studied so far. No detailed study has been carried out on the effect of sampling

disturbance and small strain behaviour of Singapore marine clay.

1.2) Objective and Scope of Work

As mentioned above, the stress-strain behaviour of deposits covering the Sin-

gapore Island is relatively unknown to date. On the other hand, there is an increasing

need for better soil constitutive models to be used in analyses for reliable estimates of

ground movements, so as to minimise cost of construction in heavily congested built

up areas. The aim of this project is to investigate the engineering properties of Singa-

pore marine clay of the Kallang formation consistent with its in-situ behaviour. This

work was conducted in close collaboration with the Port and Harbour Research Insti-

tute, Japan.

In this study, a number of aspects of the Singapore Lower Marine Clay will be

investigated using mainly undisturbed samples, though occasionally remoulded sam-

ples were also used in this study. In-situ tests including seismic cone test (SCPT),

cone penetration test (CPT), dilatometer test (DMT), field vane shear test were carried

out on a selected site at the Esplanade Park in between Connaught Drive and Raffles

Avenue, Singapore that is the proposed site for the Singapore National Arts Centre.

7
Laboratory tests were carried out in the geotechnical laboratory in The National Uni-

versity of Singapore and at the Port and Harbour Research Institute (PHRI) Japan.

Particular attention is paid to the study of the compressibility, shear strength, stiffness

and the yielding behaviour of the Singapore marine clay. With this knowledge, the in-

situ behaviour of the local clay will be compared to that of other clays, such as Ariake

and Bothkennar clays. Salient features will be singled out and evaluated. The non-

linear behaviour of local clay at low strain level will be described. Finally, the effect

of sampling disturbance on the behaviour of local clay will also be discussed.

8
Fig. 1.1 Geological map of Singapore Island (after Pitts, 1984b)

Fig. 1.2. Typical soil profile of Kallang Formation

9
 Fig. 1.3. Curves of the sea level change in the post-glacial period. (after Pitts, 1983)

Fig. 1.4. Consolidation critical stress and overconsolidation ratio versus effective
overburden stress and ground elevation, Singapore clay. (after Hanzawa and Adachi,
1983)

10