ACCEPTED MANUSCRIPT

3D Geological Modelling and Geotechnical Characteristics of

Phnom Penh Subsoils in Cambodia
Samphors Touch1, Suched Likitlersuang2* and Thirapong Pipatpongsa3

T
1
Ph.D. Student, Department of International Development Engineering, Tokyo Institute of

P
Technology, Tokyo, Japan. (Former Master Student, Department of Civil Engineering,
Faculty of Engineering, Chulalongkorn University, Bangkok, Thailand)

RI
Email: touch.s.aa@m.titech.ac.jp
2
Associate Professor, Department of Civil Engineering, Faculty of Engineering,

SC
Chulalongkorn University, Bangkok, Thailand.
Email: fceslk@eng.chula.ac.th *(Corresponding author)
3
Associate Professor, Global Scientific Information and Computing Centre, Tokyo Institute of
Technology, Tokyo, Japan

NU
Email: pthira@gsic.titech.ac.jp

Abstract: Subsoils characterisation is an important task in both geotechnical and geological

MA
engineering. In recent years, a number of subsurface investigations have increased as a result
of infrastructure construction in Phnom Penh City in Cambodia. Although there had been
investigations on the geotechnical characteristics of the subsoils, understanding soil
characteristics and 3D modelling of geological structures of Phnom Penh subsoils have not
ED

yet been tackled. This research aims to conduct a 3D geological modelling by means of
commercial software called Groundwater Modelling System (GMS), as well as provide an
insight into the geotechnical properties of Phnom Penh subsoils. Over 1,200 soil boring log
data were employed to build the 3D soil stratigraphic system in Phnom Penh City. The
PT

engineering properties of the Phnom Penh subsoils such as physical properties (i.e., grain size
distribution and Atterberg limits) and engineering properties (i.e., compressibility, undrained
shear strength and internal friction angle) have been highlighted in this paper. Finally, the
CE

analysis results of geotechnical data can be used to construct some useful empirical
correlations such as the relationship of the undrained shear strength to the SPT-N value.
AC

Keywords: geological modelling; geotechnical properties; in situ test; undrained shear

strength; Phnom Penh subsoils

Word count: 6167

Number of Table: 5

Number of Figure: 16
 ACCEPTED MANUSCRIPT

1. Introduction

Subsoil characterisation is of major importance for both geotechnical and geological

engineering involved in earthworks, structure foundations, mineral and hydrocarbon
exploration, ground water modelling, prediction and understanding of natural hazards, and
remediation of environmental issues. Studies on the geotechnical properties and the three-

T
dimensional (3D) geological structures of the subsoils have drawn much attention in recent
decades (Akpokodje, 1987; Akpokodje, 1989; Dassargues et al., 1991; de Rienzo et al., 2008;

P
Hettiarachchi and Brown, 2009; Jones and Wright, 1993; Lemon and Jones, 2003; Tonini et

RI
al., 2008).
With regard to geotechnical properties, two important considerations to be scrutinised
are whether construction will cause excessive soil deformation and/or instability due to shear

SC
failure. Thus, an understanding of the compressibility and the shear strength behaviours of
soils is an important factor in geotechnical analysis and design. To handle these matters, an
extensive number of in situ tests such as field vane shear test (FV), standard penetration test

NU
(SPT) and cone penetration test (CPT), as well as laboratory tests including consolidation test,
unconfined compression test, and direct shear box test (DSB) have been developed.
Additionally, studies on subsurface stratigraphy using the framework of 3D geological
MA
modelling have attracted much attention from many researchers in recent decades to
overcome the complex nature of the ground subsurface. With advances in computer
technology, visualisation of 3D geological structures has become an emerging research
direction, which fosters information visualisation into the geoscientific field of research
(Gallerini and De Donatis, 2009; Martti et al., 1983; Wu et al., 2005). Many 3D solid models
ED

have been proposed to elucidate the geometrical representation of geological structures, and
perform the visual analysis of spatial inhomogeneity for geological properties (Lemon and
Jones, 2003; Thierry et al., 2009; Zhu et al., 2012). A comprehensive introduction to the
PT

computer representation of complex geological objects was given by Akpokodje (1987) to

handle 3D seismic ray-tracing and velocity inversion problems. Several modelling
approaches were suggested to simulate stratified geological objects generated from different
CE

types of data such as boreholes, cross-sections, contours, and geological maps (Jones and
Wright, 1993; Lemon and Jones, 2003; Ming et al., 2010; 2012) went on to examine a 3D
geological solid modelling and presented a new approach coupling with missing strata for
AC

sedimentary stratigraphic systems.

There have been increasing subsurface investigations as a result of infrastructure
construction in Phnom Penh City in recent years. Although investigations on the geotechnical
characteristics of subsoils have been carried out by several government and non-government
sectors, understanding and modelling 3D geological structures have not yet been tackled.
Therefore, the main contributions of this study is to conduct 3D geological modelling by
means of commercial software called Groundwater Modelling System (GMS) and to provide
an insight into the geotechnical properties of Phnom Penh subsoils (Touch, 2011).
Moreover, geological and geotechnical data from over 1,200 boreholes were utilised
to construct 3D soil stratigraphic system in Phnom Penh City, Cambodia. The horizon method,
suggested by Lemon & Jones (2003), was selected for this study due to its simplicity and
efficiency in building solid models directly from boreholes with minimal user intervention.
The engineering properties of Phnom Penh subsoils such as index properties, compressibility
and shear strength behaviour have been highlighted in this study. Therefore, this work might
be the first 3D geological and geotechnical database which comprehensively presents
subsoils conditions of Phnom Penh City.
 ACCEPTED MANUSCRIPT

2. Topographic and Geological Conditions

Phnom Penh City is a capital of great socio-economic importance to the Kingdom of

Cambodia, which was founded in 1431, immediately after the Khmer kings abandoned
Angkor (Van, 2003). It is technically situated in the south-central region at the confluence of
a great lake named Tonle Sap, the Mekong River and Bassac River. From the northwest and

T
northeast, respectively, flow the Tonle Sap and Mekong Rivers, which merge and split into
the Bassac River and the Mekong. These rivers flow continuously to the southeast of South

P
China Sea. According to the available digital map provided by the Japanese International

RI
Corporation Agency (JICA) and the Ministry of Public Works and Transport in 2003 as
presented in Fig. 1, Phnom Penh City is divided into seven districts, which cover 374.43 km2.
Phnom Penh City constitutes an extensive typically flat topography. In the western part, it

SC
reaches a height of approximately 20 m above sea level inland and gently slopes continuously
toward the east to a minimum of 10 m above sea level at the west end area of the city. The
farther eastern areas including the city centre and left bank areas of the Mekong River, the

NU
Tonle Sap River and the Bassac River are mostly flat with altitude from nearly 9 m to 11 m
above sea level (JICA, 2001). Phnom Penh City remains very vulnerable to flooding because
this city was originally established on the high riverbanks and extended into lower plains
MA
lying behind the riverbanks, which are actually below flood levels. These low-lying areas
have been protected by the creation of successive concentric dikes (Van, 2003). According to
the available 1:400,000 scale from the geological map of JICA and the Ministry of Public
Works and Transport presented in Fig. 2, most areas are associated with alluvial plain and
terrace alluvial deposits in the Quaternary Period, which created loose sediment deposits
ED

from flood waters such as silt, sand and clay. Organic deposits (swamps) are also found in the
north-eastern part of Phnom Penh City.
PT

3. Three-dimensional Geological Modelling

3.1 Methodology
CE

Nowadays, 3D viewports are being used for planning purposes in geotechnical engineering.
Computers have been developed to store and output geotechnical data from the mid-1970s
onwards. This database contains detailed drilling, sampling and measurement information
AC

and 3D topology of soil boreholes (Vhaho, 1998). To meet the requirements of 3D

geological modelling, borehole-to-solid algorithm has been selected in this study via the
Groundwater Modelling System (GMS) software. This subsurface modelling method also
employed digital administrative maps and boring log data. ArcGIS software was used to
prepare the shape files and boundaries of geological modelling. These maps were then
imported and processed by GMS software. Using a conceptual model, GMS can obtain data
from ArcGIS data by mapping any important data or information for modelling procedure.

3.2 Data Initialisation

Seven districts (Chamkarmon, Prampi Makara, Dangkao, Meanchey, Russey Keo, Tuol Kork
and Daun Penh) were selected according to Phnom Penh municipalities as presented in Fig. 1.
All validated data and information on the geology, groundwater and geotechnical properties
of both superficial and subsurface were collected from relevant government and private
sectors. The data types and formats can be classified into two groups: digital maps and
hardcopy boring log data. The major source of geological information was collected from
geological survey records. Several kinds of maps were digitalised into ArcGIS for terrain
processing and reproduced according to the requirements of this research study. AutoCAD
 ACCEPTED MANUSCRIPT

software was used to read locations of boreholes for georeferencing.

3.3 Soil Boring Log Reports

Geological and geotechnical data from various construction sites should be standardised,
structured and well archived for them to be properly employed through a suitable system. The

T
data are used for efficient construction management, especially in urban areas because of
some limitations resulting from urbanisation (Chang and Park, 2004). Fig. 3 illustrates the

P
distribution of the boreholes over Phnom Penh City. Most of the boreholes are located in the

RI
eastern part. The majority of soil boring log data are located at the downtown areas
(Chamkarmon, Prampi Makara, Tuol Kork and Daun Penh), where very few investigations
have been carried out in the urban areas (Dangkao, Meanchey and Russey Keo). As presented

SC
in the land use map in Fig. 4, land use in the surrounding urban area is not yet intensive and is
mainly for agricultural purposes. Table 1 describes the classification of most Phnom Penh
subsoils used in this study. Only suitable soil boring log data within the area of interest were

NU
chosen for analyses. The selection of boreholes was based on specific criteria such as reliable
location, detailed descriptions, and geotechnical tests (field and laboratory) before being
archived in the geo-database. Soil identifications (IDs) were assigned for each material and
were recorded in the program database of GMS. Ten different soil IDs were used in this study.
MA
Following these criteria, over 1,200 boreholes were chosen for further processing.

3.4 Horizon Method

ED

The modelling result of the sedimentary system is known as the layer cake model (Turner,
2006). However, the actual strata are in fact more complicated than the layers of a cake.
Missing strata and discontinuous surfaces are very common since stratigraphic interfaces may
PT

intersect each other in complicated areas. It is not easy to effectively and accurately construct
a geologically reasonable model directly from boreholes. The investigation of the solid
modelling approach has been studied by several researchers, which can be considered as a
CE

toolset to construct 3D geological modelling structures (Gjoystdal et al., 1985; Jones and
Wright, 1993). The solid modelling operation was firstly developed to represent 3D objects in
the CAD/CAM (Computer-Aided Design/Computer-Aided Manufacturing) industry (Braid,
AC

1975; Krouse, 1985). Three-dimensional geological modelling is often exercised to construct

solids for any spatial information systems regarding environmental and geological problems.
Lemon and Jones (2003) proposed the original horizons method, which can be used to
construct the subsurface of geological modelling directly from boreholes and additional
cross-sections data. The horizons-to-solids algorithm consists of the following six steps:
assign Horizon IDs, define the primary TIN (triangulated irregular network), interpolate
horizon elevations, intersect horizon surfaces, adjust horizon elevations and build solids.

3.5 Results of 3D Geological Modelling of Phnom Penh Subsoils

Even though numerous subsurface investigations have been carried out for different
commercial projects in Phnom Penh, no comprehensive study and attempt to centralise these
activities have been undertaken. The analysis of boring log data can be used to produce a
typical geological profile as well as the representative geotechnical parameters. The resulting
solid built is shown in Fig. 5. The superficial soil distribution in the entire region is mainly
covered by made ground approximately 1.5 to 2 m thick. The representative cross-sections of
Phnom Penh subsoils are illustrated in Fig. 6. Several cross-sections have been cut for deep
understanding about geological condition of Phnom Penh subsoils. 2D cross-sections from
the north-to-south and the west-to-east were presented in the entire Phnom Penh City as
 ACCEPTED MANUSCRIPT

shown in Fig. 6(a). The representative multiple 3D geological cross-sections of Phnom Penh
subsoils associated with numerous borehole logs used in this study are exhibited in Fig. 6(b).
Based on the 3-D cross-sections, the stratigraphic framework can be categorised into three
different subsoils conditions located at the west, centre and east regions. Dangkao district is
located west of Phnom Penh and the subsoils are produced by the existing old soil.
Underneath the made ground, medium stiff to very stiff clay approximately 6-7 m thick is

T
presented. The soil layer is followed by medium to very dense sand up to a level of 15 m. On
the other hand, underground conditions at the centre of Phnom Penh (i.e., Chamkarmon,

P
Prampi Makara, Tuol Kork and Daun Penh districts) were already altered by human activities.

RI
The details of soil layers and its properties are discussed in the next section as shown in Fig.
8. In addition, river deposit creates the subsurface of the eastern areas (i.e., Meanchey and
Russey Keo districts). Very soft clay and silt layer are found from the upper layer to a depth

SC
of 12 m. Below the soft soil layer, the subsoils become medium dense to very dense sands up
to a depth of 25 m. The study of geotechnical properties has been mainly focused on the city
centre, where most site investigations for construction activities are concentrated. Therefore,

NU
the subsoils condition in downtown area is used for geotechnical properties analyses in this
research.
MA
4. In Situ and Laboratory Testing Data
4.1 In situ tests

Three in situ tests have been carried out to determine the undrained shear strength of fine-
grained soils: the standard penetration test (SPT), the pocket vane shear (PVS) test and the
ED

pocket penetration (PP) test. The SPT (ASTM D1586, 1999) is one of the most common in
situ tests used for geotechnical design in Cambodia and all around the world. Here, the SPT
was conducted by driving a standard split spoon sampler into the ground using an automatic
PT

drop hammer. In addition, the PVS test allows rapid evaluation of the undrained shear
strength of cohesive soils. The blades of the vane were pressed into the cohesive soils, and
the knob was turned slowly until the soils fail. The pocket penetration test was performed to
CE

evaluate compressive strength by pushing a 6.4-mm-diameter loading piston into the soils
being measured up to a calibration groove machined on the piston 6.4 mm from the end.
Two types of samples (undisturbed and disturbed) were taken from the field as
AC

follows:
Undisturbed samples were taken in the natural state of the soil from firm to stiff clay
and sandy clay. The undisturbed samples were taken by thin wall tube sampler in the
natural state.
Disturbed samples were taken at a rate of 1.5 m and all strata in each borehole. The
disturbed samples were also collected in soft to stiff clay and sandy soil.

4.2 Laboratory Tests

Apart from the aforementioned in situ tests, the soil index properties such as sieve analysis
and Atterberg limits were performed. For sieve analysis, a granular sample was poured into
the top sieve, which had the largest screen openings. Each lower sieve in the column had
smaller openings than the one above. Furthermore, Atterberg limits (liquid limit and plastic
limit) were also tested (ASTM D4318-10, 1998). The plastic limit test was done by rolling
out a thread of the fine portion of a soil. The water content level at which it takes 25 drops of
the cup to cause the groove to close over a distance of 13.5 mm was defined as the liquid
limit. Fig. 7a and b show typical test results of the study.
Regarding engineering properties, the unconfined compression (UC) test has been
 ACCEPTED MANUSCRIPT

carried out to determine the undrained shear strength of cohesive soils. The undrained shear
strength (su) can be determined as half of the unconfined compressive strength (qu) as shown
in Fig. 7a.
In addition, the oedometer test (ASTM D2435, 1999) was carried out to find the pre-
consolidation stress and compressibility of the cohesive soil. The consolidation time for each
loading step was selected after the sample reached the end of the primary consolidation stage.

T
5. Engineering Properties of Phnom Penh Subsoils

P
5.1 Typical subsoils profile of Phnom Penh

RI
Fig. 8 presents a typical soil profile of Phnom Penh City in downtown area up to a depth of
40 m. The testing result of the bulk unit weight, natural water content, Atterberg limits and

SC
SPT N-value are also presented in Fig. 8. It can be seen that the upper layer of Phnom Penh
subsoils consists of (i) made ground approximately 2 m thick followed by a fairly
homogeneous layer of (ii) very soft to soft clay down to a depth of 12 m and then (iii)

NU
medium stiff clay down to 17 m depth. Succeeding, six consecutive inter-bedded layers of
(iv) very loose to loose sand, (v) stiff to very stiff clay, (vi) hard silt, (vii) medium dense sand,
(viii) stiff to very stiff clay and (ix) dense to very dense sand are underneath the upper thick
MA
clay layers. The bedrock locates at approximately 40 m below ground. The Phnom Penh
geological subsoils condition is a young alluvium in the Quaternary Period and it is formed
by flood deposit.
The bulk unit weight of the Phnom Penh subsoils in Fig. 8 reflects the gradual
increases with depth from 18.44 kN/m3 to 20 kN/m3. The natural water content appears to
ED

decrease with depth from 30.54% to 13.89%. The water content is comparable to the liquid
limit at a shallow depth, which is fairly constant at about 35.39% and similar to the plastic
limit at the profound depth with an average value of 15.02%. It can be seen from Fig. 8 that
PT

most of the soils have the liquidity index (LI) between 0 to 1, which is indicative of a plastic
state (Budhu, 2011).
CE

5.2 Physical Properties

5.2.1 Grain size distribution curve
AC

The particle size distribution curves of coarse-grained soils in Phnom Penh City are depicted
in Fig. 9 and described in Table 2. The sieve analysis results of very loose to loose, medium
dense and dense, to very dense sand layers from seven districts (Chamkarmon, Prampi
Makara, Dangkao, Meanchey, Russey Keo, Tuol Kork and Daun Penh) as denoted by seven
different symbols are presented in Figs. 9a c, respectively. The results indicate that the
layers of coarse-grained soils in Phnom Penh City are composed mainly of sand particles as
almost all soil particles pass sieve no. 4 with a mesh opening of 4.75 mm. The particle size
distribution curves of medium dense sand layer from the seven districts in Fig. 8b exhibit a
marked scattering in comparison with the two other layers in Figs. 9a and c. This indicates a
difference in particle sizes. The mean particle sizes vary from 0.12 mm to 0.37 mm for very
loose to loose sand, from 0.10 mm to 0.65 mm for medium dense sand and 0.40 mm to 0.65
mm for dense to very dense sand. Moreover, Figs. 9a and b reflect a large amount of similar
fine content having particle size less than 75 m up to 42.5% for the layer of very loose to
loose and medium dense sands, while the layer of dense to very dense sand as shown in Fig.
9c consists of the maximum fine content of 27.7%, which is 14.8% considerably lower than
its counterparts.
 ACCEPTED MANUSCRIPT

5.2.2 Atterberg Limits

Fig. 10 shows a typical plot of plasticity index (PI) and liquid limit (LL) of an extensive
number of fine-grained soil samples taken at different depths up to 40 m from seven districts
(Chamkarmon, Prampi Makara, Dangkao, Meanchey, Russey Keo, Tuol Kork and Daun
Penh) in Phnom Penh City. The A-line represents a line delineating the boundaries between

T
clays (above the line) and silts and organics (below the line), while the U-line denotes a
limiting line above which PI-LL data of any soils cannot fall. The test results summary of the

P
Atterberg limits of fine-grained soils in Phnom Penh City is described in Table 3. The boring

RI
log data as shown in Figs. 10a and 10b indicate that the Phnom Penh clay at the depth up to
17 m exhibits similar plasticity index-liquid limit behaviour. It comprises of the values of LL
varying from 19.2% to 61.4% and those of PI from 2.3% to 33%. Thus, the Phnom Penh clay

SC
at shallow depth can be deemed to have low to high plasticity. Most of the PI-LL data of the
samples tend to lie close to the A-line. Figs. 10a and 10b reflect the influence of LL on the PI.
The PI increases with increasing LL. Different PI-LL behaviours for stiff to very stiff clay at

NU
the deeper depth can be seen in Fig. 10c. All the test data in Fig. 10c remain in the narrow
range of LL from 15.5% to 32.4% in comparison with those in Figs. 10a and 10b. The clay is
of low plasticity due to LL being less than 40%. In addition to the clay properties, the test
MA
data as plotted in Fig. 10d reveal that the silt comprises low to high plasticity.

5.3 Compressibility

Fig. 11 illustrates the compressibility of Phnom Penh clay up to 10 m depth. The in situ and
ED

past maximum vertical effective stress ( 'v and 'p), overconsolidation ratio (OCR), and
compression and recompression index (Cc and Cr) are plotted against depth as shown in Figs.
11ac, respectively. The in situ vertical effective stress ( 'v) at any depth in Fig. 11a was
PT

calculated as effective unit weight multiplied by corresponding depth, while the past
maximum vertical effective stress ( 'p) was computed from the analysis of the oedometer test
as depicted in Fig. 11a. Fig. 11a clearly indicates that the 'p are largely scattered and
CE

considerably higher than the 'v at shallow depth less than 5 m before reaching a similar
value at a deeper depth.
The overconsolidation ratio (OCR) as shown in Fig. 11b, which is a geotechnical
AC

parameter in relation to historical changes in the state of stress in the subsoils (Casagrande,
1936), was expressed as the ratio of 'p versus ( 'v). The OCR data appear to form a unique
curve. The OCR decreases with increasing depth. The trend of the OCR seems independent
of depth at profound depth ranging from 8 m to 10 m as the OCR value is approximately
constant. This implies that the Phnom Penh clay is subjected to overconsolidation at shallow
depth and to normally consolidated clay at deeper depth. In addition, Fig. 11c shows the
compression and recompression indices evaluated from the consolidation test as illustrated in
Fig. 7b. The compression and recompression indices (Cc and Cr) are the slope of the normal
consolidation line (NCL) and unloading line in a plot of the logarithm of vertical effective
stress against void ratio, respectively. The test data of Cc are likely to increase with depth. At
depth up to 10 m, Cc varies from 0.102 to 0.355 and Cr from 0.012 to 0.043 following by the
average value of 0.161 and 0.024, respectively.

5.4 Shear strength

5.4.1 Undrained shear strength

Estimates of the undrained shear strength of Phnom Penh clay (su) based on laboratory and
field tests including unconfined compression (UC) test, the pocket vane shear (PVS) test and
 ACCEPTED MANUSCRIPT

the pocket penetration (PP) test are illustrated in Figs. 1214, respectively. The undrained
shear strength (su) in Figs. 12b14b and the undrained shear strength normalised by in situ
vertical effective stress (su/ 'v) in Figs. 12c14c were plotted against depth. In spite of some
scattering, the test results in Figs. 12b14b obviously indicate that the su increases with depth.
As described in Table 4, the su/ 'v varies from 0.224 to 0.403 with an average of 0.29 for the
UC tests. The PVS and PP tests give somewhat lower values, about 0.21 and 0.25,

T
respectively.

P
5.4.2 Drained shear strength parameters

RI
The direct shear box test carried out under plane strain condition in which the failure plane is

SC
defined on the horizontal direction has been practically used in many geotechnical
engineering. The main advantages of this test are that the sample preparation and testing
procedure are simple. The shearing resistance from the direct shear test can be expressed
based on Mohr-Coulomb failure criterion as:

NU
c ' 'v tan ' (1)
where 'v, ' and c' are the vertical effective stress, the effective internal friction angle of soil
and the effective cohesion, respectively. The shear stress at peak state was considered as a
MA
failure state.
Fig. 15 shows the test results of effective cohesions and effective internal friction
angles obtained from the direct shear test on intact samples taken at different sites with depth
up to 10 meters. The clay specimens were trimmed to the dimension of 6 cm diameter and 2
ED

cm height by means of a cylindrical cutting ring and a wire saw. Each pair of c and was
determined from a set of test in which four different vertical effective stresses of 50, 100, 200
and 300 kPa were applied. The shearing rate of 0.01 mm/min was employed. The samples
PT

were sheared as the sample reached the end of the primary consolidation stage. The results in
Fig. 15a and b indicate that the average value of c and are approximately 37 kN/m2 and
21o, respectively.
CE

5.5. Comparison of compressibility and strength properties

AC

Table 5 shows a summary of geotechnical properties of Phnom Penh clay being compared
with Bangkok clay, Singapore clay, Hong Kong marine clay and Macau marine clay at depth
up to 10 m. These soils consist of the same Quaternary residual deposits, which originated
from sedimentation. In accordance with the test result in Table 5, Phnom Penh clay exhibits
much lower compressibility than Bangkok clay, Singapore clay, Hong Kong marine clay and
Macau marine clay. However, the overconsolidation ratios of Phnom Penh clay, Bangkok clay,
and Singapore clay are quite similar. In regard to the undrained strength properties, the
average su/ 'v of Phnom Penh clay is equal to 0.25, which is 32% less than that of Bangkok
clay, 24% higher than that of Singapore clay, 20% higher than that of Hong Kong marine clay
and 12% higher than that of Macau marine clay. In addition, Phnom Penh clay and Bangkok
reflect the similar internal friction angle of 37o and 38o while Singapore reveals lower friction
angle of 20o. As for the cohesion properties, Table 5 indicates that Phnom Penh clay,
Bangkok clay, Singapore clay, Hong Kong marine clay and Macau marine clay have a
comparable cohesion value.

6. Analysis of Undrained Shear Strength Data

The undrained shear strengths of the Phnom Penh fine-grained subsoils, which were obtained
 ACCEPTED MANUSCRIPT

from both in situ and laboratory tests, were investigated in the present study. Fig. 16 depicts
the correlation between the undrained shear strength (su) and SPT N-value. The su measured
from the pocket vane shear (PVS) test, the pocket penetration (PP) test, and the unconfined
compression (UC) test were plotted versus the corresponding number of blows for the last 0.3
m penetration of an SPT sampler. The rectangular, circular and triangular symbols represent
the su from the PVS test, the PP test, and the UC test, respectively. The test data for fine-

T
grained soils in Phnom Penh City as illustrated in Fig. 16 were screened amongst 1,200
boreholes for data quality. With regard to the in situ tests, the PP test was carried out on soft

P
to very stiff fine-grained soils while the PVS test was performed on soft soils due to the test

RI
limitations. As for the soil laboratory test, the UC test was conducted on medium to very stiff
soils. The soils having su larger than 200 kPa were removed owing to the uncertainty and
unreliability of the data.

SC
Fig. 16 shows scattered data as the su measured from the PVS test, the PP test, and the
UC test were correlated with the SPT-N value. It could be that the effects of unit weight,
liquid limit, plasticity index, confining pressure and initial degree of saturation, which were

NU
not taken into account in the present study, have influences on the su response to some
degrees (Kong and Yang, 1998; Nassaji and Kalantari, 2011; Sivrikaya, 2009; Sowers, 1979;
Stroud, 1974). Additionally, Fig. 16 indicates that almost all test data of su and the SPT-N
MA
value will apparently fall within a range formed by the approximate upper (UB) and lower
(LB) bounds by dash dot lines, which can be respectively expressed in the linear forms as
follows:
17N (UB)
su (kN/m2 ) (2)
ED

3N (LB)
The test data from the foregoing types of testing were fitted by means of the least square
linear regression as shown by the solid line in Fig. 16. Thus, the unique empirical correlation
PT

between su measured from three different types of tests and the SPT-N value for the particular
fine-grained subsoils in Phnom Penh City leads to
su (kN/m2 ) 5.5N (3)
CE

It is of interest to compare this correlation other common correlation proposed by previous

investigators. Terzaghi and Peck (1967), performing the extensive in-situ and field tests on
fine-grained soils, suggested the relationship of su and the SPT-N value.
AC

su (kN/m2 ) 6.5N (4)

The empirical correlations as expressed in Eqs.(3) and (4) prove that the su evaluated from
the SPT-N value in the present study is 15% less than that of the study proposed by Terzaghi
and Peck (1967); hence providing more conservative value.
In addition, Sambhandharaksa and Taesiri (1987) investigated the correlation of su
with SPT-N value for Bangkok clay as follows:
For high plastic clay
su (kN/m2 ) 5.20N (5)
For low plastic clay
su (kN/m2 ) 6.82N (6)
Eq (3) indicates that the su of Phnom Penh clay is slightly higher than that of the high plastic
Bangkok clay as expressed in Eq. (5) and is considerably lower than that of the low plastic
Bangkok clay in Eq. (6). Besides, the su of Phonom Penh clay is also a bit higher in
comparison with Singapore clay with the su equal to 5N (Tang et al., 2004).

7. Conclusion
 ACCEPTED MANUSCRIPT

The analysis results are useful for future research and study on subsurface condition in
Phnom Penh as well as civil engineering practices. This study is considered as the first
research document on Phnom Penh subsoils. Three-dimensional geological modelling and
geotechnical characteristics of Phnom Penh subsoils were investigated in this study. The main
conclusions can be summarised as follows:
1) The Phnom Penh subsoils consist of thick layer of clay up to 17 m depth, following six

T
consecutive inter-bedded layers.
2) The layers of coarse-grained soils in Phnom Penh City are composed mainly of sand

P
particles with the finer content up to 42.5%.

RI
3) The Phnom Penh clay at shallow depth can be considered as low to medium plasticity
clay.
4) The Phnom Penh clay undergoes lightly overconsolidation at shallow depth and normally

SC
consolidated clay at deeper depth.
5) Beneath 10 m depth, the normalised undrained shear strength varies from 0.21 to 0.29.
As for the drained shear strength parameters, the Phnom Penh clay consists of cohesion

NU
and friction angle equal to 36.8 kN/m2 and 21.3o, respectively.
6) The empirical correlation was proposed to evaluate the undrained shear strength of the
Phnom Penh clay from the SPT-N value.
MA
Acknowledgement

This work was supported by the AUN/SEED-Net (JICA) under the collaborative research the
ED

collaborative research program 20102011. The first author appreciatively acknowledges the
scholarship of the master and doctoral degree programs provided by JICA. The authors would
like to thank all technical staffs at the Research and Design Enterprise (RDE) for their
kindness in encouraging and providing relevant data for carrying out academic research
PT

activities related to such important works.

References
CE

Akpokodje, E.G. 1987. The engineering-geological characteristics and classification of the

major superficial soils of the Niger Delta. Engineering Geology, 23, 193-211, doi:
AC

10.1016/0013-7952(87)90090-1.
Akpokodje, E.G. 1989. Preliminary studies on the geotechnical characteristics of the Niger
delta subsoils. Engineering Geology, 26, 247-259, doi: 10.1016/0013-7952(89)90012-4.
ASTM D1586. 1999. Standard test method for penetration test and split-barrel sampling of
soils. 1999 ed. ASTM International, West Conshohocken, PA.
ASTM D2435. 1999. Standard test method for one-dimensional consolidation properties of
soils.
ASTM D4318-10. 1998. Standard test methods for liquid limit, plastic limit, and plasticity
index of soils.
Balasubramaniam, A.S., Hwang, Z.M., Waheed, U., Chaudhry, A.R. & Li, Y.G. 1978. Critical
state parameter sand peak stress envelopes for Bangkok clays. Quarterly Journal of
Engineering Geology, 1, 219-232.
Bergado, D.T., Chai, J.C., Miura, N. & Balasubramaniam, A.S. 1998. PVD improvement of
soft Bangkok clay with combined vacuum and reduced sand embankment preloading.
Geotechnical Engineering, Southeast Asian Geotechnical Society, 29, 95-121.
Braid, I.C. 1975. The synthesis of solids bounded by many faces. Commun. ACM, 18, 209-
216, doi: 10.1145/360715.360727.
Casagrande, A. 1936. The determination of the preconsolidation load and its practice
 ACCEPTED MANUSCRIPT

significance. Proc. 1st International Conference on Soil Mechanics and Foundation

Engineering, Cambridge, Mass, 60-64.
Chang, Y.-S. & Park, H.-D. 2004. Development of a web-based Geographic Information
System for the management of borehole and geological data. Computers & Geosciences,
30, 887-897.
Dassargues, A., Biver, P. & Monjoie, A. 1991. Geotechnical properties of the Quaternary

T
sediments in Shanghai. Engineering Geology, 31, 71-90, doi: 10.1016/0013-
7952(91)90058-s.

P
de Rienzo, F., Oreste, P. & Pelizza, S. 2008. Subsurface geological-geotechnical modelling to

RI
sustain underground civil planning. Engineering Geology, 96, 187-204, doi:
http://dx.doi.org/10.1016/j.enggeo.2007.11.002.
Gallerini, G. & De Donatis, M. 2009. 3D modeling using geognostic data: The case of the

SC
low valley of Foglia river (Italy). Computers & Geosciences, 35, 146-164.
Gjoystdal, H., Reinhardsen, J.E. & AsteboL, K. 1985. Computer Representation of Complex
3-D Geological Structures using a New "Solid Modeling" Technique. Geophysical

NU
Prospecting, 33, 1195-1211, doi: 10.1111/j.1365-2478.1985.tb01359.x.
Hettiarachchi, H. & Brown, T. 2009. Use of SPT blow counts to estimate shear strength
properties of soils: Energy balance approach. Journal of Geotechnical and
MA
Geoenvironmental Engineering, 135, 830-834.
JICA. 2001. Basic design study report on the flood protection and drainage improvement
project in the municipality of Phnom Penh in the Kingdom of Cambodia.
Jones, N.L. & Wright, S.G. 1993. Subsurface characterization with solid models. Journal of
ED

Geotechnieal Engineering, 119, 1823-1839.

Kong, S.K. & Yang, D.Q. 1998. Shear strength characterstics of residual soils from
sedimentary Jurong formation in Singapore. 13th Southeast Asian Geotechnical
Conference, Taipei, Taiwan.
PT

Krouse, J.K. 1985. Solid modeling catches on. Machine Design, 57, 6064.
Lemon, A.M. & Jones, N.L. 2003. Building solid models from boreholes and user-defined
cross-sections. Computers & Geosciences, 29, 547-555.
CE

Lee, K.M., Ng, P.C.C. (1999). A geotechnical investigation of marine deposits in a nearshore
seabed for land reclamation. Canadian Geotechnical Journal 36(6), 981-1000.
Likitlersuang, S., Teachavorasinskun, S., Surarak, C., Oh, E. & Balasubramaniam, A. 2013.
AC

Small strain stiffness and stiffness degradation curve of Bangkok Clays. Soils and
Foundations, 53, 498-509, doi: http://dx.doi.org/10.1016/j.sandf.2013.06.003.
Martti, M., ntyl & Markku, T. 1983. Localized set operations for solid modeling. SIGGRAPH
Comput. Graph., 17, 279-288, doi: 10.1145/964967.801159.
Ming, J., Pan, M., Qu, H. & Ge, Z. 2010. GSIS: A 3D geological multi-body modeling
system from netty cross-sections with topology. Computers & Geosciences, 36, 756-767.
Nassaji, F. & Kalantari, B. 2011. SPT capability to estimate undrained shear strength of fine-
grained soils of Tehran, Iran. EJGE, 16, 1229-1238.
Sambhandharaksa, S. & Taesiri, T. 1987. development of theory and prictice in geotechnical
engineering. the 8th ARC on SMFE. Thailand Department of Highways, Kyoto ,Japan.
Sivrikaya, O. 2009. Comparison of Artificial Neural Networks models with correlative works
on undrained shear strength. Eurasian Soil Science, 42, 1487-1496, doi:
10.1134/s1064229309130092.
Sowers, G.F. 1979. Introductory soil mechanics and foundations. 4th ed. Macmillan, New
York.
Stroud, M.A. 1974. The standard penetration test in insensitive clays and soft rock. the 1st
European Symposium on Penetration Testing. Stockholm, Sweden, 367-375.
Tanaka, H., Locat, J., Shibuya, S., Soon, T.T. & Shiwakoti, D.R. 2001. Characterization of
 ACCEPTED MANUSCRIPT

Singapore, Bangkok, and Ariake clays. Canadian Geotechnical Journal, 38, 378-400.
Tang, S.K., Tong, S.Y., Lim, T.L., Sugawara, S. & Furusono, T. 2004. Assessment of Ground
Condition for a Mass Rapid Transit System in Singapore. Proceedings of World Tunnel
Congress and 13th ITA Assembly, Singapore, 22-27.
Terzaghi, K. & Peck, R.B. 1967. Soil mechanics in Engineering Practice. John Wiley, New
York.

T
Thierry, P., Prunier-Leparmentier, A.-M., Lembezat, C., Vanoudheusden, E. & Vernoux, J.-F.
2009. 3D geological modelling at urban scale and mapping of ground movement

P
susceptibility from gypsum dissolution: The Paris example (France). Engineering

RI
Geology, 105, 51-64.
Tonini, A., Guastaldi, E., Massa, G. & Conti, P. 2008. 3D geo-mapping based on surface data
for preliminary study of underground works: A case study in Val Topina (Central Italy).

SC
Engineering Geology, 99, 61-69.
Touch, S. 2011. Three-dimensional geological modeling of Phnom Penh subsoils. Master
Thesis, Department of Civil Engineering, Faculty of Engineering, Chulalongkorn

NU
University.
Turner, A.K. 2006. Challenges and trends for geological modelling and visualisation. Bulletin
of Engineering Geology and the Environment, 65, 109-127, doi: 10.1007/s10064-005-
MA
0015-0.
Vhaho, I. 1998. From geotechnical maps to three-dimensional models. Tunnelling and
Underground Space Technology, 13, 51-56, doi: 10.1016/s0886-7798(98)00020-0.
Van, M. 2003. Modern Khmer Cities. Reyum, Phnom Penh.
ED

Wu, Q., Xu, H. & Zou, X. 2005. An effective method for 3D geological modeling with multi-
source data integration. Computers & Geosciences, 31, 35-43.
Yan, W.M., Ma, Yongfeng. (2010). Geotechnical characterization of Macau marine deposits.
Engineering Geology, 113(1-4), 62-69.
PT

Zhu, L., Zhang, C., Li, M., Pan, X. & Sun, J. 2012. Building 3D solid models of sedimentary
stratigraphic systems from borehole data: An automatic method and case studies.
Engineering Geology, 127, 1-13.
CE
AC
 ACCEPTED MANUSCRIPT

List of Figures

P T
RI
SC
NU
MA
Figure 1: Phnom Penh Municipality districts map (reproduced from JICA & Ministry of
Public Works and Transport, 2003)
ED
PT
CE
AC

Figure 2: Geological map of Phnom Penh City (reproduced from JICA & Ministry of Public
Works and Transport, 2003)
 ACCEPTED MANUSCRIPT

P T
RI
SC
NU
Figure 3: Locations of over 1,200 boreholes used in the study
MA
ED
PT
CE
AC

Figure 4: Land use map in 2003 (reproduced from JICA & Ministry of Public Works and
Transport, 2003)
 ACCEPTED MANUSCRIPT

T
Made ground

P
Very soft to Soft clay
Stiff to very stiff clay

RI
Dense to very dense sand
Hard silt
Soft silt

SC
Organic soil
Medium dense sand
Medium stiff clay
Very loose to loose sand

NU
Figure 5: 3D geological model of Phnom Penh subsoils in oblique view with vertical
exaggerated coefficient equals 100
MA
(a)
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

(b)

T
Made ground

P
Very soft to Soft clay
Stiff to very stiff clay

RI
Z Dense to very dense sand
Y Hard silt
Soft silt

SC
Organic soil
X Medium dense sand
Medium stiff clay

NU
Very loose to loose sand
Figure 6: Geological cross-sections of Phnom Penh subsoils. (a) Multiple cross-sections in
plan view and (b) multiple cross-sections in oblique view with vertical exaggerated
coefficient equals 100
MA
ED
PT
CE
AC

Figure 7: Laboratory tests on Phnom Penh stiff lean clay sample taken at the depth of 26.5 m
below the ground surface. (a) Unconfined compression test and (b) 1D consolidation test
 ACCEPTED MANUSCRIPT

P T
RI
SC
NU
MA
Figure 8: Typical subsoil profile of Phnom Penh City in downtown area
ED
PT
CE
AC

Figure 9: Particle size distribution curves of coarse-grained soils for all seven districts of
Phnom Penh subsoils. (a) Very loose to loose sand, (b) Medium dense sand and (c) Dense to
very dense sand
 ACCEPTED MANUSCRIPT

P T
RI
SC
NU
MA

Figure 10: Correlations between plasticity index and liquid limit of Phnom Penh subsoils. (a)
ED

Very soft to soft clay, (b) Medium stiff clay, (c) Stiff to very stiff clay and (d) Hard silt
PT
CE
AC

Figure 11: Compressibility of Phnom Penh clay. (a) In situ vertical effective stress ( v ) and
maximum past stress ( p ) versus depth, (b) Overconsolidation ratio (OCR) versus depth and
(c) Compression index (Cc) and recompression index (Cr) versus depth
 ACCEPTED MANUSCRIPT

P T
RI
SC
NU
MA
Figure 12: Undrained shear strength from unconfined compression (UC) test. (a) Phnom
Penh soil profile, (b) Undrained shear strength versus depth and (c) Normalised undrained
shear strength versus depth
ED
PT
CE
AC

Figure 13: Undrained shear strength from pocket vane shear (PVS) test. (a) Phnom Penh soil
profile, (b) Undrained shear strength versus depth and (c) Normalised undrained shear
strength versus depth
 ACCEPTED MANUSCRIPT

P T
RI
SC
NU
Figure 14: Undrained shear strength from pocket penetration (PP) test. (a) Phnom Penh soil
MA
profile, (b) Undrained shear strength versus depth and (c) Normalised undrained shear
strength versus depth
ED
PT
CE
AC

Figure 15: (a) Cohesion and (b) angle of internal friction from direct shear box test
 ACCEPTED MANUSCRIPT

P T
RI
SC
NU
Figure 16: Empirical correlations of the undrained shear strength to the SPT N-value of
Phnom Penh fine-grained soils
MA
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

List of Tables
Table 1 Classification of most Phnom Penh subsoils
Soil Type Soil ID
Made ground Made ground 1
Dense to very dense sand 2
Sand Medium dense sand 3

T
Very loose to loose sand 4

P
Stiff to very stiff clay 5

RI
Clay Medium stiff clay 6
Very soft to soft clay 7

SC
Hard silt 8
Silt
Soft silt 9
Organic Organic soil 10

Very loose toNU

Table 2: Particle size distribution of coarse-grained soils in Phnom Penh City
Particle size distribution Medium dense Dense to very dense
MA
parameters loose sand sand sand
D10 (mm) 0.0391 0.0223 0.037
D50 (mm) 0.223 0.245 0.481
Cu 6.879 15.262 19.077
ED

Cc 1.861 0.834 1.611

USCS symbol SW SP SW
Note: D10 (mm): Effective size, D50 (mm): median size, Cu (coefficient of uniformity) =
PT

D60/D10, Cc (coefficient of curvature) = D302/ (D60D10)

Table 3: Atterberg limits of fine-grained soils in Phnom Penh City

CE

Descript Very soft to medium Stiff to very stiff

stiff clays clays
Natural water content, wn (%) 26.9 19.4
AC

Liquid limit, LL (%) 36.5 39.3

Plastic limit, PL (%) 14 17.4
Plasticity index, PI (%) 22.5 21.9
Liquidity index, LI (%) 57.3 9.1

Table 4: Normalised undrained shear strength parameters of Phnom Penh clay from UC, PVS
and PP tests
Unconfined
Pocket Vane Shear Pocket Penetration
Tests Compression (UC)
(PVS) test (PP) test
test
su v 0.29 0.21 0.25
 ACCEPTED MANUSCRIPT

Table 5 Summary of geotechnical engineering properties of Phnom Penh clay, Bangkok clay, Singapore clay, Hong Kong marine clay, and
Macau marine clay at depth up to 10 m
Parameter Phnom Penh Bangkok Singapore Hong Kong Macau marine
Soil properties

T
s clay clay clay marine clay clay

IP
0.102 - 0.355 0.691 -
Cc 0.65 - 0.75e 0.308 - 1.315f 0.41g

CR
Compressibility (0.161) 1.682a
and stress 0.012 - 0.043 0.069 - 0.0198 -
Cr - 0.08g
history (0.024) 0.184 a 0.1241f

US
OCR 1.5 1.3b 1.1 - 1.4b - -
Normalised

N
undrained shear su v 0.21 - 0.29 (0.25) 0.33c 0.19b 0.20f 0.28g

MA
strength
Drained shear c (kN/m2) 37 38d 20e - -
strength ( )
o 21 24 d 25 e - 27 32g

D
a b c
Bergado et al. (1998); Tanaka et al. (2001); Likitlersuang et al. (2013);

TE
d
Balasubramaniam et al. (1978); e Tang et al. (2004); fLee and Ng (1999); and gYan and Ma (2010).

P
CE
AC

23
 ACCEPTED MANUSCRIPT

Highlights:

The work presents a 3D geological model using GMS software.

The analysis results employ over 1,200 boring log data.
The engineering properties of the Phnom Penh subsoil are summarised.
This paper could be considered as the first research document on Phnom Penh subsoil.

P T
RI
SC
NU
MA
ED
PT
CE
AC

24