KSCE Journal of Civil Engineering (0000) 00(0):1-14 Geotechnical Engineering

Copyright ⓒ2015 Korean Society of Civil Engineers

DOI 10.1007/s12205-015-0470-5 pISSN 1226-7988, eISSN 1976-3808
www.springer.com/12205
TECHNICAL NOTE

Practice and Experience in Deep Excavations in Soft Soil

of Ho Chi Minh City, Vietnam
Nguyen Kiet Hung* and N. Phienwej**
Received August 20, 2014/Revised 1st: May 6, 2015, 2nd: August 22, 2015/Accepted September 22, 2015/Published Online November 18, 2015

··································································································································································································································

Abstract

Deep excavations for the construction of basements and urban infrastructure have increased drastically in Vietnam. This study
reviews 18 cases of deep excavations in 4-16-m-thick soft clay in central Ho Chi Minh City (HCMC), three of which experienced
severe instability. Most reviewed cases applied internally braced Diaphragm Walls (DWs) whose maximum lateral wall deflections
were 0.15%-1.0% of the excavation depth. In cases of a sheet pile wall or micro-bored pile wall, this value ranged 1.0%-2.4%.
Among wall types used in HCMC, the DW was most effective in minimizing wall and ground movement. The wall deflection
magnitude in cantilever mode at the first stage of excavation reached 35%-60% of total wall deflection in the final stage. Increasing
the stiffness of the DW and bracing system is shown to be ineffective. The case studies suggest a need to: i) formulate local
guidelines or codes of practice in association with damage assessment, and ii) improve numerical analysis in the assessment of wall
system design, ground movements and risk of damage to adjacent buildings. Furthermore, for numerical analyses using the non-
linear hardening soil model of PLAXIS, more realistic soil parameters for HCMC soft clay are explored in laboratory testing.
Keywords: Ho Chi Minh City, deep excavations, soft clay, case history, hardening soil model, PLAXIS
··································································································································································································································

1. Introduction HCMC. Section 3 reviews and assesses current local practices

used in the design and construction of deep excavations.
Ho Chi Minh City (HCMC) has emerged as a pillar of recent Improvements in design practice are suggested, with the
economic development within the Southern Focal Economic improvement of numerical analyses highlighted in Section 4. In
Region in Vietnam. Government has invested and will continue section 5, by conducting a laboratory testing program, a realistic
to invest greatly in upgrading the utilities and transportation soil modulus and strength parameters for Finite Element (FE)
infrastructure of HCMC. A rail-based Mass Rapid Transit analysis of deep excavations in HCMC soft ground using the
(MRT) system for HCMC, having six lines with a total length of hardening soil model are explored and suggested. Finally, section
about 195.3 km (this includes approximately 80 km of underground 6 present conclusions drawn from the study and provides
sections), has been planned. MRT lines 1 and 2 are now under recommendations.
construction; the deepest excavation is 40 m. In recent years,
many deep basements were constructed for high-rise buildings 2. Soft Ground Zone in HCMC
servicing as car parks and shops. These excavations were as deep
as 24 m from the surface. Deep excavations in soft ground in 2.1 Geologic Setting of HCMC
populated urban areas always involve high risk. In recent years, The geology of HCMC has characteristics of a bordering
there have been a number of publicized cases of failures of wall region in transition, from the rising mass of the Dalat midlands
systems used in deep excavation projects in central HCMC. It is of central Vietnam to the lowland depression of the Mekong
thus necessary to assess weaknesses and areas for improvement Deltas and characteristics of a transmission from low mound
in past and current design and construction practices in an effort land remaining from the Neogene to deltaic lowland, resulting in
to formulate necessary local guidelines or codes of practice for the powerful movement of surface flows from west to east
use in future projects. during the Holocene. Igneous and metamorphic bedrocks are
The remainder of the present paper is organized as follows. found at depths of 130-200 m. The subsoil profile to a depth of
Section 2 reviews geological and geohydrological conditions of 70-80 m can be described by 10 strata (Figs. 1 and 2).
the subsoils and geotechnical properties of soft ground in (i) Made ground: Fill soil, mostly 0.5-2.0 m thick.

*Doctoral Student, School of Civil Engineering, Asian Institute of Technology, P.O. Box 4, Klong Luang, Pathumthani 12120, Thailand (Corresponding
Author, E-mail: st109062@ait.asia, nguyenkiet_hung@yahoo.com)
**Professor, School of Civil Engineering, Asian Institute of Technology, P.O. Box 4, Klong Luang, Pathumthani 12120, Thailand (E-mail: noppadol54@gmail.com)

−1−
 Nguyen Kiet Hung and N. Phienwej

part of the city adjacent to the Saigon River mouth, and having
surface elevation below 2 m. The stratum thickness is mostly 2-
16 m, but can be greater than 30 m in the south. The soil has high
water content; wn = 81.62% on average. The unconfined
compressive strength qu ranges 20.3-49.1 kPa. The SPT N value
is lower than 5.
(iii) Stratum 2: Loose fine sand with a limited distribution in
the south with a lens form and thickness ranging 1-10 m and
sometimes higher. SPT N ranges 5-10.
(iv) Stratum 3: Medium to stiff sandy clay having outcrops in
the north and center of the city and thickness of 10-40 m. wn =
6.3%-38.4% and qu = 27-67 kPa.
(v) Stratum 4: Loose to medium dense sand with gravel,
encountered in northern and central areas at a depth of
approximately 10 m. SPT N ranges 7-30.
(vi) Stratum 5: Stiff to very stiff sandy clay, distributed mostly
in the north and northeast of the city, with thickness ranging 3-20
Fig. 1. HCMC Engineering Geological Map (recompiled from Pham,
2008) m and SPT N ranging 9-30.
(vii) Stratum 6: Medium dense sand with a limited distribution,
with thickness ranging 5-8 m and an SPT N value of about 26.
(ii) Stratum 1: Soft to very soft organic clay covering 63% of (viii) Stratum 7: Medium to stiff sandy clay encountered at a
the city including the central lowlands and the entire southern depth of 10-35 m, only in the center and south of the city, with

Fig. 2. Typical Geotechnical Sections of the City Center Lowlands (recompiled from Pham, 2008)

Fig. 3. Hydrogeological Section Q-Q' (after DEHGSV, 2006)

−2− KSCE Journal of Civil Engineering

 Practice and Experience in Deep Excavations in Soft Soil of Ho Chi Minh City, Vietnam

thickness ranging 20-30 m and qu ranging 80-450 kPa.

(ix) Stratum 8: Fairly dense medium sand located at a depth of
10-40 m, ending at a depth of 30-60 m. SPT N ranges from 15 to
more than 30.
(x) Stratum 9: Stiff to very stiff sandy clay with thickness
ranging 1.5-30 m at a depth of 30-65 m. qu ranges 80-250 kPa.
(xi) Stratum 10: Dense, fine to coarse sand covering almost the
whole HCMC area, with thickness of 20 to more than 40 m, at a
depth of 40-70 m, with SPT N ranging 30-100.
There are five aquifer layers in the HCMC area (DHEGSV,
2006), as illustrated in Fig. 3. The Holocene and Pleistocene
Fig. 4. Soft-ground Distribution in the HCMC Central Lowlands
aquifers are the two aquifers that may directly affect deep (Ha et al., 2006)
excavation works. There are also the Upper Pliocene aquifer,
Lower Pliocene aquifer and Mesozoic aquifer. The overlying
Holocene aquifer is located in the city lowlands along the Saigon 1, as described above) with thickness of 4-25 m underlying the 1-
River valley and has thickness ranging from some meters to 42 2-m-thick made ground is mainly river-marine swampy (bog)
m. The aquifer includes multi-genesis sediments (alluvial, alluvial- sediment from the transgression in the late Holocene of the Can
marine and alluvial-marine-swampy sediments) that are made up Gio formation overlying Pleistocene and Pliocene sediments.
of muddy clay and silty clay mixed with fine sand and it is Fig. 5 shows the subsoil profile along MTR Line 1, a large-scale
occasionally embedded with a lens of fine sand with humus. The infrastructure project situated within the zone as mentioned in
perched groundwater level fluctuates at depths of 0-2 m. The Section 1.
Pleistocene aquifer is located widely under the city and exposed
in the central and northern districts of the city at a depth of 40-80 2.2 General Geotechnical Characteristics of the HCMC
m. The aquifer contains fine to medium and coarse sand, Central Soft Ground Zone Through a 70-m Depth Profile
occasionally mixed with gravel. Regarding subsoil conditions in the central city pertinent to
The city center lowland zone is located adjacent to the Saigon deep excavations, this study derives a generalized subsoil profile
River. Fig. 4 is a map of the soft ground distribution and and soil properties on the basis of soil investigation data of
thickness in HCMC (Ha et al., 2007). The soft clay layer (stratum selected large-scale infrastructure and high-rise building projects

Fig. 5. General Soil Properties Along the Ben Thanh-Suoi Tien Metro Line (MRT line 1)

Fig. 6. Generalized Soil Properties of the HCMC Central Soft Ground Zone Through a 70-m Depth Profile

Vol. 00, No. 0 / 000 0000 −3−

 Nguyen Kiet Hung and N. Phienwej

Table 1. Basic Properties (average) of HCMC Soft Clay in the City permeable sandy strata (strata 2 and 4) has been the primary
Central Lowlands cause of major cases of failures during deep excavations in
Natural water content, wn 80% HCMC in recent years.
Liquid limit, LL 89%
Plastic limit, PL 36% 3. Practice in Deep Excavations
Plasticity index, PI 53
Unit weight, γ 15 kN/m3 3.1 Reviewed Cases
Specific gravity, Gs 2.68 There have been more than 50 projects of deep excavations for
Clay content 63% the building of basements and infrastructure development in the
HCMC lowlands. These excavations have used an earth retaining
system comprising concrete diaphragm walls (DWs), contiguous
in the area. Fig. 6 presents the generalized soil profile together Bored Pile Walls (BPWs) and steel sheet pile walls (SPWs) with
with indices and strength properties. The dominant soil strata in internal bracing. Excavations in 18 projects on soft ground with
the area are (i) made ground, (ii) stratum 1: soft to very soft an excavation depth ranging 7-24 m for at least two basement
organic clay, (iii) stratum 2: loose fine sand (in lens form, not floors, for which design and measurement data were largely
encountered in many places), (iv) stratum 3: medium to stiff clay available, were selected as case studies in this research. There are
(not encountered in some places), (v) stratum 4: fine sand—a two popular methods, top-down and bottom-up methods, for the
loose to medium dense stratum that may be encountered between excavation and concreting of basement structures. The most
the above strata, (vi) stratum 7: a medium to stiff clay stratum in commonly used wall type is the DW. Owing to the presence of
lens form, as the Pleistocene aquifer, (vii) stratum 8: fairly dense soft clay and land right regulations, only the internal bracing
medium sand, as the Upper Pliocene aquifer, and (viii) stratum system could be used for lateral support of the walls. The
10: dense, fine to coarse sand. excavation depths ranged 8-24 m with the majority being greater
Typical engineering properties of the soft to very soft organic than 12 m and four being greater than 19 m. Four projects had
clay, hereinafter called HCMC soft clay, are given in Table 1. It large excavation areas with length exceeding 150 m. SPWs were
is noted that, no matter how cautiously the deep excavations used in two excavations having depths shallower than 12 m. The
were made, with the presence of soft to very soft organic clay thicknesses of DWs and BPWs were 0.80-1.0 m for excavations
(stratum 1) throughout the central lowlands of HCMC, there are deeper than 12 m, except in the case of the 1.20-m-thick DW in
major concerns about excessive lateral wall displacement and the the deepest excavation. BPWs with diameters of 0.30 and 0.40 m
associated ground settlement, which are often the primary cause were used in two excavations with depths of 8.0 and 11.0 m,
of damage to adjacent structures. In addition, the presence of the while a DW with thickness of 0.60 m was used in an 11-m-deep

Table 2. Basic Data of 18 Deep Excavations in HCMC Central Soft Clay

Retaining wall data
Preload value
δ max
Cases Method Hw H t Number of δmax --------- B×L ∆H1 of bracing
Type
bracings H
m m m mm % m×m m kN/m
A TD DW 37 19.4 0.8 5 28 0.15 67 × 78 1.5 0
B BU DW 40 20 1.2 7 32 0.16 22 × 57 1.8 50; 70; 200; 300
C TD DW 54 13.5 0.8 2 78 0.50 80 × 160 1.5 0
D TD DW 21 14.5 0.8 2 50 0.30 68 × 165 1.0 0
E BU SPW 18 7 0.4 2 172 2.40 120 × 170 1.5
F BU DW 22 11 0.6 2 33 0.30 47 × 56 1.6 150
G BU BPW 18 11 0.4 2 120 1.00 35 × 96 2 120
H BU DW 26.5 14 1.0 4 38 0.27 59 × 64 1.5 50; 75
I BU BPW 12 8 0.3 2 158 1.90 60 × 150
J BU SPW 24 12 0.4 2 118 1.00 56 × 65 1.5 50
K TD DW 24.5 14 0.8 4 80 0.50 50 × 58.3 2.5 0
L TD DW 38.5 15 0.8 4 94 0.63 45 1.5
M TD DW 49.1 24 1.2 6 210 0.88 80 × 91 1.5 0
N TD DW 31 16 0.8 3 12 0.50 95 × 140 0
O TD DW 25 14 0.8 3 65 × 83 0
P TD DW 45 20 1.0 5 200 1.0 40 × 44 1.0 0
Q BU DW 39 14.8 0.8 4 68 0.46 30 × 60 1.5
R TD DW 38 16.5 1.0 4 33 0.20 52 × 63 1.5 0

−4− KSCE Journal of Civil Engineering

 Practice and Experience in Deep Excavations in Soft Soil of Ho Chi Minh City, Vietnam

excavation. For excavations deeper than 12 m, walls with a deep triaxial compression test; UC: Unconfined compression test;
toe embedment depth (i.e., a ratio of wall length to excavation UU: Unconsolidated undrained triaxial compression test; DSS:
depth of 1.75 to 2.65) were adopted. The deepest excavation of Direct simple shear test; OED: Oedometer test; BH: Borehole for
24 m was made with DWs having length of 49 m. The vertical soil investigation. (vi) A: Saigon MC tower—D1; B: Thi Nghe
spacing of the bracing of DWs in the excavations ranged 3.50- pump station—BT; C: Sunrise City—Plot V-D7; D: Cantavil
7.50 m; walls were braced or the top-down method was used for Complex—D2; E: Blooming Park Towers—D2; F: Thanh Da
support while walls were not tied back. Soil investigations view—BT; G: Saigon Pearl—BT; H: Saigon Times squares—
conducted for the design of the piled foundation and excavations D1; I: Phu My Thuan—NB; J: Daewon-Hoancau Building—
of these projects are summarized in Table 2. Instrumentation BT; K: Petroland Tower—D7; L: Le Meriden Saigon—D1; M:
implemented for safety monitoring of the excavations is also Vincom B towers—D1; N: Kumho plaza—D1; O: Bitexco
summarized. Currently, there are about 10 capable foundation Financial Tower—D1; P: Pacific Building—D3; Q: Lim
contractors, both international and local, providing services for Tower—D1; R: Vietcombank Tower—D1.
the construction of DWs and BPWs in HCMC.
Notes: (i) The ground water level is at a depth of 1.5 m (on 3.1.1 Wall Movement
average) in all cases (except for 8 m in Case P). Performance of deep excavations in urban areas may be
(ii) TD: Top-down; BU: Bottom-up; DW: Diaphragm wall; gauged by the level of induced wall and ground movements. The
BPW: Micro bored pile wall; SPW: Sheet pile wall; Hw: Wall present study analyzed data on lateral wall movements that were
depth; H: Excavation depth; t: Wall thickness; δmax: Maximum commonly monitored by inclinometers during staged excavations
wall horizontal deflection; B: Excavation width; L: Excavation of projects. Settlements of ground surface in areas next to the
length; ∆H1: Free length from the top of the wall to the level of excavation pit could hardly be monitored because of obstruction
the first excavation in cantilever mode of the retaining wall; D1, by, for example, existing buildings. Figs. 7 and 8 show the
D2, D7, BT, NB: Districts 1, 2, 7, Binh Thanh, Nha Be. (iii) relationship between the maximum magnitude of lateral wall
Underlined values 200, 1.0: δmax, δmax/H computed employing deflection monitored by an inclinometer δmax and excavation
the FEM. (iv) In: Inclinometer measurement; S: Settlement depth H and the relationship between δmax/H and H for 17
marker measurement; Tm: Tilt meter measurement; Sg: Strain excavations in the HCMC central lowlands. There is no increase
gauge measurement; Pz: Piezometer measurement; O: Observation in δmax with H, which suggests that δmax is affected not only by H,
well. (v) SPT: Standard penetration test; CU: Consolidated but also by other factors such as the wall type, the wall and
undrained triaxial compression test; CD: Consolidated drained support system stiffness and the excavation method. Larger wall

Table 3. Summary of Soil Tests, Instrumentation and Effects on Adjacent Buildings in the Reviewed Cases
Types of main soil Observations Damage to the Failure occurrence with
Case
tests conducted conducted adjacent Type/Cause
Failure occurrence due to DW structure,
A SPT, CU, DSS, OED, pump test In, S, Tm, Sg, Pz, O Extremely severe
ground water
B SPT, DSS, OED In, S, Pz, O No damage No
C SPT, CU, DSS, OED In, Pz, O Some damage No
D SPT, UU on soft clay, DSS, OED In, S, Pz, O Some damage No
SPW fails in bending;
E SPT, DSS In Some damage
severe instability occurs.
F SPT, CU & UU, DSS, OED In, S No damage No
G SPT, DSS, OED In Severe damage Instability in wall, support
Failure incident due to DW structure,
H SPT, DSS, OED In, S, Tm, Sg, Pz, O, C Severe damage
ground water.
I SPT, DSS In Severe damage Instability in wall and support
J SPT, DSS In Some damage Bracing strut fails by buckling.
K SPT, CU, UU, VST, OED In, S, Pz, O Some damage No
L SPT, CU, DSS In, Pz, O Some damage No
M SPT, CU, DSS, OED In, Pz, O Some damage No
N SPT, DSS, OED In, Pz, O No damage No
O SPT, CU, DSS, OED No damage No
SPT, DSS. Failure due to DW structure, ground water.
P No measurement Extremely severe
1BH × 80 m; 1BH × 45 m Basal heave stability is at margin level.
CU & UU for stiff soils, DSS, CD, Failure incident due to DW structure,
Q In, Pz, O Severe damage
OED, SPT ground water.
R SPT, DSS No No

Vol. 00, No. 0 / 000 0000 −5−

 Nguyen Kiet Hung and N. Phienwej

1% for the DW are compared with cases in the literature as

follows. (i) According to Manna and Clough (1981), with a low
FSheave for an SPW in soft clay, δmax/H may reach 2%, while for
DW, δmax/H may reduce to 0.5%. (ii) Clough et al. (1989) drew a
diagram for the estimation of δmax in soft clay depending on
FSheave and system stiffness and showed that δmax/H may reduce
to about 0.2% for the DW with high FSheave (>3.0) and high
support system stiffness (>300) and may reach 3% in cases of the
SPW with low FSheave (<1.0) and low support system stiffness
(<50). (iii) Long (2001) investigated 296 cases of excavation in
soft clay where there is transition from soft to stiff clay below the
bottom of the cuts and found that, in the cases of a propped wall
or top-down construction method, normalized lateral movement
mostly ranged from 0.1% to 1% of excavation depth; meanwhile,
Fig. 7. δmax of Wall Versus Excavation Depth Compared with Moor- large movements of up to δmax/H = 3.2% might occur for soft
mann’s (2004) Range of δmax = 0.5-1% H (0.87% on aver-
clays with low FSheave. Results similar to those of Long (2001)
age)
are obtained in the cases of the DW in the present study as seen
in Fig. 8; meanwhile, higher values are obtained for the SPW or
BPW (from 1% to 2.5%, approximately). This similarity can be
explained by the similarity in the transition from soft to stiff
clays occurring below the excavation bottom of the cases
studied. (iv) Moormann (2004) conducted an extensive empirical
study on 530 case histories of deep excavation in soft clay (cu<
75 kPa) and found that δmax varies from 0.5% to 1% of H and has
an average value of 0.87% of H. δmax/H is less than 0.9% for the
DW when using a braced-wall support with H < 22 m, ranges
from 0.1% to 0.75% for the DW when using the top-down
method with H < 22 m, and can exceed 1% for the SPW and
soldier pile wall.
The average δmax in the present study is 0.61% of H and within
Fig. 8. δmax/H Versus Excavation Depth Compared with the Results of the range of Moormann (2004) as shown in Fig. 7. δmax/H in the
Long (2001) in Cases of Propped Walls/top-down Method cases using the DW is low, and similar to that in all cases
reported in the above studies. However, when using an SPW or
movements occurred in cases that an SPW or BPW was used, BPW, owing to low FSheave and low support system stiffness,
where the excavation bottom was overlying the soft soil stratum. δmaxvalues are similar to the ranges of Manna and Clough (1981),
The two cases with the deepest excavation depths were cases M Clough et al. (1989), and Long (2001) but higher than the upper
and P (Table 2). Excavation in case P experienced severe basal limit of 1% of H of Moormann (2004).
heave instability due to poor structure and ground water control. The effect of wall movement on the damage response of the
Even though the project employed a DW with a top-down adjacent structures during excavation in HCMC central soft
construction method, the wall displacement was greater than that ground has not yet been studied in the literature. Ou (2006)
in the cases that an SPW or BPW was used. Larger wall stated that the factors causing wall deformation will also produce
displacements were due to a low factor of safety against basal ground settlement. Previously, Goldberg et al. (1976) analyzed
heave (FSheave) and low wall flexural stiffness (Figs. 11, 12, 13, 63 case histories and found that, for soft to stiff clays, the factor
and 14). BPWs that were used in HCMC for deep excavation δmax/δv,max may lie from 0.75 to above 2.0. The studies of Manna
had mostly small piles, ranging 300-400 mm in diameter and and Clough (1981), Ou (1993) and Hsieh and Ou (1998)
having low FSheave and support system stiffness, and only used concluded that δv,max = (0.5-0.75) δmax; as for very soft soil, δv,max
with the bottom-up method. Larger-diameter piles were rarely might equal or exceed δmax. According to Moormann’s (2004)
used for walls because of their higher cost. Another disadvantage empirical relationship between δmax and δv,max depending on the
of the BPW is that, for deep excavations in such soft ground, type of soil, retaining wall and support system used, the factor
displacement of the wall is relatively large (e.g., cases G and I). δmax/δv,max for soft clay may range 0.3-2.0 for an SPW with braced
In the situation of high ground water pressure, leakage through support and δmax < 12 cm, 0.75-1.5 for an SPW with braced
walls may occur in contrast to the case of DWs. support and δmax > 12 cm, 0.5-1.0 for a DW with braced support
The cases for HCMC investigated in the present study with and δmax > 12 cm, and 0.5-2.0 for a DW with δmax < 12 cm. δv,max
δmax/H ranging 1% to 2.4% for the SPW or BPW and 0.15% to is therefore directly proportional to δmax.

−6− KSCE Journal of Civil Engineering

 Practice and Experience in Deep Excavations in Soft Soil of Ho Chi Minh City, Vietnam

Among the various criteria for damage categorization available

in the literature, the method of Boscardin and Cording (1989)
based on the building response described in terms of horizontal
strain εh and angular distortion β was recommended by Tan and
Chow (2008). Meanwhile, Mair et al. (1996) proposed a state-of-
the-art procedure with which to compute the limiting tensile
strain (εh, β) induced in the building from the green-field
excavation-induced ground displacement projected on the building.
The level of damage is then classified according to strain level. It
is seen from the literature (e.g., Son and Cording, 2005) that the
magnitude of green-field excavation-induced ground displacement
(δv,max), which relates to δmax, is only one of many factors affecting Fig. 9. Observed Wall Displacement-case B
the level of damage; e.g., the building stiffness, building weight,
interface of building and soil, and type of building foundation. It
can thus be said that low values of wall displacement, which
result in small β and consequently small εh (Boscardin and
Cording, 1989), indirectly indicate low levels of damage to
adjacent buildings. However, high values of wall displacement
might not suggest high levels of damage. Moh et al. (1999)
emphasized that among potential preventive measures that can
reduce ground movements (i.e., reducing ground movements at
sources, restricting ground movements from propagating,
strengthening structures and correcting building movements), the
most effective measures involve reducing ground movement at
sources, including wall deflection and consolidation of soft clay. Fig. 10. Observed Wall Displacement-case K
However, reducing the consolidation of soft clay in congested
HCMC urban areas is infeasible because of the high cost and the
time required. The most effective and feasible measure, with ∆H1 was 12 mm/32 mm (case B), 37 mm/78 mm (case C), 58
which to reduce ground movement, is reducing wall deflection. mm/172 mm (case E), 40 mm/120 mm (case G), 14 mm/38 mm
Therefore, the most effective measure that can be undertaken to (case H), 56 mm/158 mm (case I), 30 mm/80 mm (case K), 35
mitigate damage to adjacent buildings is to reduce wall mm/94 mm (case L), and 27 mm/68 mm (case Q). It can be said
deflection. that, in HCMC soft clay, minimizing δmax1 is important in
It is thus seen that using a DW rather than other wall types in reducing cumulative δmax, and therefore, as discussed above, in
deep excavations in HCMC soft ground is the most effective mitigating damage to adjacent buildings. The location of the first
measure for reducing wall movement, ground movement and bracing should be carefully assessed and selected to optimize the
damage to adjacent buildings. wall performance. Bowles (1997) noted that, in the case of clay,
the depth of first excavation should not exceed the depth of
3.1.2 Relationship between H and δmax/H. potential tension cracks obtained with reasonable FS in stability
There are three types of relationship between δmax/H and H in analysis. It should be emphasized that the occurrence of cracks
the cases reviewed. (i) When using a DW with high wall support on the ground surface, which may be filled with water,
system stiffness, wall bending stiffness and FSheave, δmax/H = reduces soil strength through the softening of clay and
0.15%-0.6%. (ii) When using a DW with H deeper than 15 m consequently increases lateral pressure. In top-down construction,
with the excavation bottom level at a considerable depth in hard to be on standby for a future raised level of the ground floor
soil strata but with low wall support system stiffness, wall due to a raised flood level under an increased high-tide level
bending stiffness and FSheave, and even instability as in case P, of newly developed or renovated streets in the proximity, the
δmax/H = 0.88%-1%. (iii) When using an SPW or BPW, δmax/H = finished ground floor level of the new buildings with
1%-2.4% (Fig. 8). basements underneath is mostly 1.0-1.5 m above the existing
In most cases, the free length ∆H1, measured from the top of adjacent street level. Consequently, the first excavation depth
the retaining wall to the level of the first excavation while the of top-down constructions is to be dredged at 1.5-2.5 m so
wall was in cantilever mode in supporting the ground, ranged that the first basement floor can be firstly cast. This ∆H1 can
1.5-2.5 m. δmax1/∆H1 values of the first excavation stage varied be reduced (as in case D) if the DW top level is raised above
accordingly in the range 0.5%-2.0%. The wall movement in the ground level instead of above the first basement floor to
cantilever mode δmax1 reached approximately 35%-60% of the accommodate for the casting of the ground floor in the initial
maximum displacement after the final excavation stage; δmax1/ excavation stage.

Vol. 00, No. 0 / 000 0000 −7−

 Nguyen Kiet Hung and N. Phienwej

3.1.3 Relationship between δmax/H and FS against basal analytically deduced by Mana and Clough (1981).
heave Figures 11 and 12 show the correlation between δmax normalized
Figure 11 compares the relationship between δmax/H and FSheave by excavation depth and FSheave with reference to contour lines of
with that established for the SPW and slurry wall. The factor of FSheave established by Clough and O’Rourke (1990). It is seen
safety against basal heave defined by Terzaghi (1943) was that the cases within or at the margin of the instability zone
applied for the excavations with H < B in all the studied cases. limited by the contour line FSheave = 1 are the cases of instability
Long (2001) suggested that, in cases where the retaining wall is given in Table 3. Most of these cases involve the use of an SPW
keyed into a hard soil layer, there is high FSheave likely to be or BPW (cases E, I, J, and G) and the only case using the DW is
higher than 3.0. The ratio δmax/H in the cases of using the DW case P, where basal heave failure occurred as described in
does not vary in accordance with FSheave; i.e., there is no Section 3.2.
relationship between either the maximum wall displacement or
ground movement and FSheave in the excavations using the DW. 3.1.4 Normalized Maximum Lateral Wall Movement Ver-
This behavior of excavations in HCMC soils is similar to that sus System Stiffness
reported in a study on Bangkok soft clay (Phienwej and Gan, Figure 13 shows the relationship between δmax/H and EI/(γwl4)
2003). This indicates that, when using a DW for deep excavations with reference to the contour lines of FSheave proposed by Fernie
in HCMC soft clay, the evaluation or estimation of wall lateral and Suckling (1996). The relationship implies that, in the case of
displacement or ground movement is independent of FSheave. using a DW for deep excavations in soft ground, FS of safety
However, the limit δmax/H = 0.5% can be established; i.e., δmax/H against basal heave of soil below the base of excavation does not
≥ 0.5% can be estimated when FSheave < 2 while δmax/H < 0.5% appear to be a relevant parameter in assessment of lateral wall
can be applied when FSheave ≥ 2. This limit for the case studies is deflection. Ulrich (1989) studied and presented the average trend
in accordance with the relationship between FSheave and δmax/H line of the correlation between wall bending stiffness EI/L3 and
δmax/H and the range of this correlation for all soils. A similar
relationship was found in the present study; however, the

Fig. 11. δmax/H Versus FS Against Basal heave

Fig. 13. δmax/H Versus System Stiffness EI/γwl (with reference to Fernie
and Suckling, 1996)

Fig. 12. δmax/H Versus System Stiffness with Contours of FS Against Fig. 14. Wall Bending Stiffness EI/L3 Versus δmax/H (with reference
Basal Heave to Ulrich, 1989)

−8− KSCE Journal of Civil Engineering

 Practice and Experience in Deep Excavations in Soft Soil of Ho Chi Minh City, Vietnam

correlation is scattered and unable to show a defined trend line

(Fig. 14). This means that increasing or decreasing wall bending
stiffness does not affect normalized lateral wall movement. The
results are supported by the findings of Ulrich (1989).

3.2 Cases with Instability Problems

3.2.1 Case P
In May 2007, a semi top-down excavation was made for a
five-level basement of a 21-story building on 1750 m2 of land in Fig. 16. Layout Plan and Section of Case A-Damage to and Col-
the city center. The 20-m-deep excavation was carried out with lapse of Adjacent Buildings in February 2010 Due to Deep
aid of a 1000-mm-thick and 45-m-deep DW. The subsurface Excavation
strata included 1-m-thick made ground overlying 4-m-thick silty
clay with c = 13 kPa and ϕ = 11o, and 4.1-m-thick stiff clay;
underneath these strata was a 29-m-thick layer of fine sand (i.e., 30-m depth to fill cavities in soil and cut off water inflow as
the aquifer described in Section 2). Water stops between DW illustrated in Fig. 15.
panels were set only from the ground surface to a depth of 22 m.
In October 2007, during the final excavation stage of casting a 3.2.2 Case A
foundation raft, a void that was 30-35 cm wide and 168 cm long A 19.8-m top-down excavation using a 0.8-1.0-m-thick and
was found in the DW at a depth of 21 m. Despite many measures 27-m-deep DW was carried out for a five-level basement of a 40-
taken to seal the gap and water leakage, severe ingress that led to story building on 4700 m2 of land in the city center. The subsoil
soil piping in the sand layer behind the wall could not be included 3-m-thick made ground and 8-m-thick soft silty clay
controlled. The ground water head at the leakage point was 18.5 overlaying 7-m-thick loose to dense sand, 3-m-thick medium
m. The basal failure occurred because the sandy soil beneath the clay, and 13.5-m-thick medium dense sand. Underneath these
dredged level under undrained conditions behaved mainly as a strata was 13.5-m-thick hard clay overlaying dense sand. At two
frictionless material. Thirty minutes after the leakage appeared, corners of the site there were 0.6-m-thick and 10-m-deep DW
piping of the sand and cavities behind the wall resulted in panels that had been previously cast for the originally planned
excessive subsidence of the ground surface and severe damage to excavation of a three-level basement. These existing DWs were
adjacent two-story building blocks, with settlement as large as incorporated with new DWs of deeper depth as an earth-
200 mm (Nguyen, 2010). The incident was followed by lawsuits. retaining structure of the excavation. An observation program
Even though complaints were filed against the project by the including inclinometers, piezometers and tilt meters was
neighborhood prior to the failure owing to ground movements organized to monitor the wall displacement, ground water level
and cracks of buildings, the contractor did not take proper and tilt of the adjacent buildings. At the end of January 2011,
measures to ensure the safety of adjacent buildings. The during the excavation of the third basement floor at an 11-m
contractor/engineer involved in the work lacked experience in depth, there was a structural failure of the DW in the adjoining
deep excavation; this was the first time they worked on this type area of the existing and new DW panels as illustrated in Fig. 16.
of construction. It appeared that no observation or instrumentation Ground water inflow at the gap between old and new DWs could
program was carried out for the excavation. The remedial not be handled by the dewatering system. There was severe
measures included adding two rows of jet grouting piles (JGPs) cracking and the tilt of nine town houses. Additionally, two
with diameters of 600-800 mm behind the DW joints from a 10- adjoining blocks of old two-story buildings collapsed. Fortunately,
residents were safely evacuated without injury following an
issuance of warning of danger from the uncontrolled water
ingress. Similar to case P, JGPs were used as a remedial measure.
A series of 1000-mm-diameter JGPs was installed at depths of 9-
18 m. The treatment was successful and the remedial excavation
was successfully completed in April 2011.

3.2.3 Case H
In 2009, a 14-m-deep excavation was carried out to construct a
three-level basement of a 37-story building using a 0.8-m-thick
and 35-m-deep DW. The subsoil strata included 1.5-2-m-thick
fill, 2.5-4-m-thick soft clay, and 10-m-thick sand. Previously, in
Fig. 15. Case P with Failure of the Diaphragm Wall Structure and 2008, a set of 0.6-mm-thick and 12-m-deep DW panels and
Ground Water Piping into the Excavation bored piles had been completed for a one-level basement only.

Vol. 00, No. 0 / 000 0000 −9−

 Nguyen Kiet Hung and N. Phienwej

Fig. 17. Case H: Serious Ground Water In-flow Due to Failure of the DW Structure and Watertightness with Damaged Adjacent Houses
and a Remedial Measure

There were then seven untreated and poor contact locations impact of deep excavations may not be directly applicable.
among the new DW panels, original bored piles and original DW
panels resulting in large infiltration of ground water and sandy 3.3 Weakness of the Current Design Practice in HCMC
soil. During the excavation, cavities formed at these locations
owing to ground water and fine sand flowing unexpectedly into 3.3.1 Subsurface Investigation and Laboratory Testing
the excavation. The gradual piping of the sand layer caused In cases of basements under high-rise buildings, soil investigations
settlement and damage to adjacent buildings. Lasting complaints, are mostly done for the design of piling and foundations, while
lawsuits and flooding inside the excavation caused delay of the determination of soil parameters for the design of deep
construction from April 2009 to March 2010, when the failure excavations is secondary (Table 3). Laboratory tests that
was completely solved. At first, the treatment was performed by determine the shear strength of soil are usually performed on
grout injection with a series of 60-mm-diameter boreholes in the disturbed and undisturbed soil samples collected from boreholes.
proximity of the DW panels and outside the excavation. The most common tests are direct shear tests and unconfined
However, this solution did not help stop the ground water inflow; compression tests. Sometimes, UU tests and CU triaxial tests are
observation revealed that cracks in four adjacent town houses conducted. CD tests are rarely done. Piezometers are rarely
continued to widen and there were continuous complaints. JGPs installed and the groundwater level is normally surveyed through
were then performed together with needle grouting to seal investigation of boreholes, which is not reliable.
leakages that appeared in the DW. Another series of grouting
injections was applied underneath the adjacent buildings during 3.3.2 Check of Excavation Basal Instability
dewatering with a pumping system and casting the foundation Owing to the presence of stratum 2 (loose fine sand) with a
and basement structures. Finally, in March 2010, after almost a lens form and stratum 4 (loose to medium sand) in the HCMC
year’s delay, the problems were resolved. The layout of the central lowlands, the present study revealed that nearly all the
excavation support system, the ground condition and the deep excavation failures in HCMC soft ground involved
treatment method performed are illustrated in Fig. 17. groundwater problems, with cases A, H, P, and Q being typical.
In cases A and P, as described above, basal heave failure
3.2.4 Cases with Severe Damages to Adjacent Buildings occurred because, during the excavation period, from the
Table 2 shows that wall displacements in many of the reviewed hydraulic failure, the sandy soil beneath the excavation bottom
excavation cases exceeded 50 mm (i.e., cases C, E, G, I, J, K, L behaved mainly as a frictionless material under undrained
and M). Intolerable impacts in the form of cracks and excessive conditions. For cases of groundwater existing in sandy soils both
settlement of buildings in the proximity of the excavations were behind the wall and below the bottom of the excavation, a
experienced, leading to complaints and delays of the works. detailed stability check of stability against hydraulic failure in the
Severe damage to adjacent buildings occurred in five of the form of the upheave of the excavation bottom and water ingress
excavations owing to excessive ground movements from and piping erosion is crucial but often has not received due
instability of the excavation or wall movements. attention in past projects. It is seen that the cases within or at the
Generally, deep excavations in soft clay are considered as margin of the instability zone limited by the contour line FSheave =
acceptable and building protection measures may not be required 1 are the cases of instability given in Table 3. Most of these cases
if wall deflection can be kept to no more than 25–50 mm (Moh involve the use of an SPW or BPW (i.e., cases E, I, J, and G)
and Hwang, 1999). However, the actual level of damage depends while the only case of using a DW is case P, where basal heave
on many other factors including the conditions and integrity of failure occurred as described in Section 3.2.1.
building structures. In HCMC, many existing buildings adjacent
to excavations of modern development projects are old, dating to 3.3.3 Method of Design Analysis
pre-war times; thus, such a general guideline on the potential The continuum finite element method of analysis has been

− 10 − KSCE Journal of Civil Engineering

 Practice and Experience in Deep Excavations in Soft Soil of Ho Chi Minh City, Vietnam

increasingly used in Vietnam in recent years. The most commonly of deep excavation in HCMC were due to groundwater, the
adopted software is PLAXIS, for which the Mohr–Coulomb unsuitability of structural members and the instability of the wall
model of soil behavior with a loading stiffness parameter is and support system. These failures might have been prevented or
commonly used. However, owing to its simplification of soil avoided if conditions had been carefully surveyed and design
behavior, the Mohr–Coulomb model has limitations and yields principles and quality control checks had been applied accordingly.
erroneous results in terms of the ground settlement distribution. Beyond a certain threshold, ground movement in excavation can
Thus, its use should be carefully supervised. Proper understanding cause the intolerable cracking of adjacent buildings; especially
and use of the constitutive soil model is a crucial requirement in those buildings that are constructed on a shallow foundation and
producing a safe and sound design. For instance, the case of the consist of brick walls (see Tables 2 and 3). As shown in Section
Nicoll Highway collapse in Singapore demonstrated the 3.1.1, the most effective measures that can be taken to mitigate
consequence of an incorrect use of the soil model in the design damage to adjacent buildings involve reducing wall deflections
of deep excavation in soft soil (Yong and Lee, 2007). Advanced during excavation. Therefore, wall displacement can act as an
models, such as the Hardening Soil Model (HSM), which are indicator in the assessment of the soundness of a design and the
more accurate for predicting non-linear deformation behavior of extent of damage caused by deep excavations. The concerned
soil in deep excavations, cannot yet be readily implemented authorities and offices should outline guidelines on the design of
owing to a lack of required soil parameter data in soil deep excavations in HCMC soft clay, which should include a
investigation practice (Table 3) with necessary calibrations. procedure with which to assess the reliability of designs, any
possible damage to an adjacent structure and compensation
3.3.4 Damage Assessment and Observational Methods agreements or lawsuit adjudications. The guidelines need to
Although serious shortcomings have been observed in the specify the allowable levels of wall displacement and ground
design and construction of deep excavations in recent years, settlement that induce limited cracking or tilting to quantify the
there have been no guidelines or regulations on damage design parameters of the deep excavation support system and
assessment/prevention and compensation for adjacent buildings. procedures. To ensure the effectiveness of such guidelines,
In absence of such guidelines, design practice in HCMC has references to Eurocode 7, NAVFAC 1982a, CIRIA C517:
addressed the ultimate limit state design rather than the serviceability Temporary propping of deep excavations—guidance of design,
limit state design, where wall and ground movement is a focal and CIRIA C580: Embedded retaining wall—guidance for
point of the design consideration. In addition, a survey of a economic design should be incorporated. A design procedure
building condition is often not conducted adequately for the that combines the ultimate limit state and serviceability limit
monitoring of deep excavation performance and quantitative state (see Chapter 6, CIRIA C580) can be referred to and
assessment of damage. Practically, a 20-mm limit of wall modified in accordance with local conditions.
displacement can be regarded as “acceptable” in the preliminary
review for the design and evaluation of protection against 4.2 Selection of Wall Type, Support and Method of Exca-
damage to adjacent structures within a distance of 3-5 m from vation
the deep excavation boundary. Observation and instrumentation The DW instead of the SPW or micro BPW (of low stiffness
practice of deep excavations in HCMC (Table 3) has normally and large displacement as described in sections 3.1.1 and 3.1.3,
been limited to the monitoring of the wall lateral displacement Figs. 8, 11, 12, 13, and 14) using top-down or bottom-up
and ground water level. Mostly, a dilapidation geodetic survey of construction is recommended for deep excavations because of its
neighboring buildings, even those buildings in poor state, has not higher stiffness with respect to ground movement controls and its
been carried out prior to the commencement of excavation. The
settlement, tilt and cracking of building in adjacent areas have
not been commonly monitored prior to the occurrence of
damage. In some cases (e.g., case P), no instrumentation at all
was installed until failure occurred.

4. Suggestions for the Improvement of Design

Practice in HCMC

4.1 Need for Proper Guidelines on Design

Owing to the presence of very soft to soft clay (stratum 1),
shallow ground water and sandy strata (strata 2 and 4) as
described in Section 2, current challenges in deep excavations in
HCMC central lowlands are to “avoid failures” and “mitigate Fig. 18. Oedometric Moduli Versus Consolidation Pressure from an
damage to adjacent buildings” (see the case studies in Section 3). Oedometer Test-sample OED1-11 on Soft Clay at 8.5 m
The review conducted in this study shows that the major failures Depth, Site 1

Vol. 00, No. 0 / 000 0000 − 11 −

 Nguyen Kiet Hung and N. Phienwej

Fig. 20. Variation of Deformation Moduli with Confining Pressure of

Soft Clay Samples at Site 1 from CID Tests

selection of appropriate soil parameters, especially the soil

Fig. 19. Stress Path of the CIU Test Series from Site 1 and Strength stiffness of soft clay, which is the most important parameter in
Parameters
computing ground movements and wall displacements, to obtain
results with the required accuracy. In each project case, even
most efficient watertightness. Top-down methods should be though CID tests are expensive, together with CIU and oedometer
applied in the city center, where minimizing ground movement tests, they need to be conducted, properly calibrated and well
for the protection of adjacent structures is the top priority. A DW back-analyzed as illustrated in the section 5 hereinafter to satisfy
can be combined with the permanent building foundation and the aforesaid requirement.
basement wall for the functions of temporary shoring, hydraulic
cut off, vertical support elements and shear walls. To ensure 5. Determination of Strength and Stiffness Para-
stability against hydraulic failure, a DW should be designed so meters of HCMC Soft Clay in Laboratory Test-
that it can be suitably embedded in the impermeable layer ing for Soil Modeling using the HSM
underneath. However, in the case of a deep layer of impervious
soil (e.g., Case H), it would be infeasible to insert a DW deeply The main issue with using the HSM in Vietnam is that E ref 50

enough to ensure the total cutoff of seepage flow. Because it can and E ref
ur moduli are difficult to determine from CD tests, which

be costly and difficult to insert a DW deeper in the sandy stratum are expensive and take a long time. In such cases, the designer
of soil, lowering the ground water table by pumping to relieve may rely on available oedometer test results. However, the main
the piezometric pressure and maintaining sufficient FS against problem with using oedometer test results to determine soil
“blow-up” is a feasible solution. Pumping tests, which have stiffness parameters for the HSM of soft clay is establishing the
rarely been done (Table 3), need to be performed prior to the correlation between the stiffness parameters derived from
design and construction. oedometer test results and those derived from CD test results:
E ur ⁄ Eur, oed and E ur, oed ⁄ Eoed . Plaxis (Brrinkgreve et al., 2010)
ref ref ref ref

4.3 Selection of Constitutive Soil Models for FE Analyses proposed the default ratios E ref ur = 3E 50 and E 50 = 1.25 E oed ,
ref ref ref

of Deep Excavations while Gebreselassie (2003) proposed that Eur ⁄ E50 and E ur, oed ⁄ Eref
ref ref ref
oed

Instead of the most popular Mohr–Coulomb model, the are 8.43 and 4.29, respectively, on average. E ref ur ⁄ E 50 is appro-
ref

advanced HSM should be used in soil modeling as part of deep ximately 10 and Eref ur, oed ⁄ E oed ranges 4-8.5 in the case of Bangkok
ref

excavation analyses in HCMC for the accurate calculation and soft clay (Surarak, 2012).
prediction of wall displacement and ground movement. The In this research, to investigate the strength and stiffness parameters
important issue for a proper analysis using the HSM is the of HCMC clays relevant to FE analyses of deep excavations, a

Table 4. Soil Parameters of HCMC Soft Clay Recommended for Soil Modeling using the HSM

Eur, oed Eur Eur

ref ref ref
E oed Eur, oed E50 Eur -------------
- -------------
- -------
-
ref ref ref ref
Depth of soft c' ϕ' m Rf
E oed Eur, oed E50
ref ref ref
clay stratum
kPa kPa kPa kPa [−] [−] [−] kPa [o] [−] [−]
665 4680 2222 12200 7.0 2.6 5.5 9 20.6 0.94 0.87
From 4 to 13 m
675 1360 8000 5.9 9 20.6 0.90 0.90
1375 9695 3200 12068 7.1 1.2 5.7 11 20.2 0.90 0.91
From 13 to 20 m
965 1800 9800 5.4 11 20.2 0.90 0.90
Notes: Bold italic numbers are soil parameters calibrated by FE simulation of the triaxial tests with best-fit lines.

− 12 − KSCE Journal of Civil Engineering

 Practice and Experience in Deep Excavations in Soft Soil of Ho Chi Minh City, Vietnam

laboratory testing program was conducted with a series of 45 References

undisturbed specimens taken from two sites. The tests performed
Boscardin, M. D. and Cording, E. J. (1989). “Building response to
include oedometer tests, constant rate strain consolidation tests
excavation-induced settlement.” Journal of Geotechnical Engineering,
with vertical drainage (CRS-V), and consolidated undrained and Vol. 115, No. 1, pp. 1-21, DOI: 10.1061/(ASCE)0733-9410(1989)
drained triaxial compression tests under both isotropic and 115:1(1).
isotropic conditions. While the details of the test program and Bowles, J. E. (1997). Foundation Analysis and Design, 5th International
findings are reported in another paper, the results are briefly Ed., McGraw-Hill, New York, N.Y, pp. 795-797.
presented here in Figs. 18-20 and Table 4. Brinkgreve, R. B. J., Swolfs, W. M., Engin E., Waterman, D., Chesaru,
A., Bonnier, P. G., and Galavi, V. (2010). PLAXIS 2D 2010-
Material models manual, pp. 50-56.
6. Conclusions
Clough, G. W. and O’Rourke, T. D. (1990). “Construction induced
movements of in-situ walls.” Design and performance of earth
The main contributions and findings of the study are summarized retaining structures, Proceedings of a Specialty Conference at
as follows. Cornell University, ASCE, New York, pp. 439-470.
1) Geotechnical characteristics of the soft ground zone of Clough, G. W., Smith, E. M., and Sweeney, B. P. (1989). “Movement
HCMC central lowlands were reviewed. The subsoils consist of control of excavation support systems by iterative design.” ASCE,
a very soft to soft clay layer covering 63% of the city and GSP, Vol. 22, No. 2, pp. 869-884.
saturated sand layers existing within and below the deep Cogeco (2011). Inclinometer monitoring report, Lim Tower project,
excavation depth range. The subsoil conditions pose potential Internal publication.
Division of Hydrogeology and Engineering Geology for South of
risk in excavation works.
Vietnam (DHEGSV) (2006). Internal publication.
2) The review and characterization of retaining walls with European Committee for Standardisation (2004). Eurocode 7: Geotechnical
excavation depth H and maximum wall lateral displacement δmax Design. Part 1: General Rules, CEN, Brussels.
in historical cases revealed the following. (i) δmax /H = 0.15%- Fernie R. and Suckling T. (1996). “Simplified approach for estimating
0.6% for a diaphragm wall with high system stiffness, wall lateral wall movement of embedded walls in UK ground.” Proc. Int.
bending stiffness and factor of safety against basal heave; δmax/H Conf. on Geotechnical Aspects of Underground Construction in Soft
= 0.88%-1% for a diaphragm wall with H deeper than 20 m and ground, London.
Gaba, A. R., Simpson, B., Powrie, W., and Beadman, D.R. (2003).
the excavation bottom at considerable depth in hard soil strata
Embedded Retaining Wall – Guidance for Economic Design (CIRIA
but with low stiffness of the wall support system, wall bending C580), CIRIA, London.
stiffness and factor of safety against basal heave, even in the case Gebreselassie, B. (2003). “Experimental, analytical and numerical
of instability; and δmax/H = 1%-2.4%, approximately, for a SPW investigations of excavations in normally consolidated soft soils.
or BPW. (ii) In the first excavation stage, δmax/H values varied Dissertation.” University of Kassel. Schriftenreihe Geotechnik, Heft,
from 0.5% to 2% and were approximately 35% to 60% of the Vol. 14, pp. 149-151.
maximum value in the final excavation stage. Goldberg, D. T., Jaworski, W. E., and Gordon, M. D. (1976). Lateral
3) Most failures in the deep excavation cases were due to Support and Underpinning, Report FHWA-RD-75-128, 1, Federal
Highway Administration, Washington D.C. (PB 257210).
instability of the design and ground water and structural failures,
Ha, P. T. S. and Son, L. M. (2007). “Applying kriging to predict the
which can be prevented by appropriate design and strict quality distribution of the holocene soft soil stratum in the urban of Ho Chi
control of workmanship and construction procedures. To prevent Minh city.” Science and Technology Development, Vol. 10, No. 2,
such failures and to protect nearby structures, the measures of pp. 43-53.
using jet grouting or a cut-off wall may be applied prior to the Hiep, N. V. (2010). Basement breakdowns in HCM city – Some
commencement of excavation. experiences, Lecture, French-Vietnamese Conference CIGOS 2010
4) There is an urgent need for legal and sound guidelines for High-rise Buildings and Underground Structures.
the design of deep excavations in HCMC soft ground, not only Hsieh, P. G. and Ou, C. Y. (1998). “Shape of ground surface settlement
profiles caused by excavation.” Canadian Geotechnical Journal,
for the design itself but also for damage assessment and damage
Vol. 35, No. 6, pp. 1004-1017.
compensation lawsuits and construction quality control. Huyndai Mobis, J. V. (2004, 2005, 2006). Sequential displacement data
5) The non-linear modulus values and strength values of – VW strain gauge record sheet–HCMC Package # 8 Pump Station,
HCMC soft clay obtained with the hardening soil model were Internal publication.
investigated in a thorough laboratory testing program and then Long Van Co., Ltd (2011). Geotechnical monitoring report – Cantavil
calibrated by FE simulation of triaxial tests. Specific values of Complex towers, Internal publication.
the modulus and strength were recommended for a better Long, M. (2001). “Database for retaining wall and ground movements due
prediction of ground movements associated with deep excavations to deep excavations.” Journal of Geotechnical and Geoenvironmental
Engineering, Vol. 127, No. 3, pp. 203-224, DOI: 10.1061/
in the city soft ground employing continuum FE analysis. Back
(ASCE)1090-0241(2001)127:3(203).
analyses for further calibration of soil parameters of HCMC soft Mair, R. J., Taylor, R. N., and Burland, J. B. (1996). Prediction of
clays are necessary for the accurate prediction of ground ground movements and assessment of risk of building damage due to
movement and wall displacement in deep excavations and bored tunneling, Int. Symp. Geotech. Aspects Underground Constr.
tunneling works in HCMC. Soft Ground. E. Mair & Taylor. London, Appril 1996, Rotterdam,

Vol. 00, No. 0 / 000 0000 − 13 −

 Nguyen Kiet Hung and N. Phienwej

Balkema. Society, Vol. 34, No. 3, pp. 167-175.

Mana, A. I. and Clough, G. W. (1981). “Prediction of movements for Son, M. and Cording, E. J. (2005). “Estimation of building damage due
braced cuts in clays.” Journal of the Geotechnical Engineering to excavation-induced ground movements.” Journal of Geotechnical
Division, Vol. 107, No. 8, pp.759-777. and Geoenvironmental Engineering, Vol. 131, No. 2, pp. 162-177,
Moh Z. C., Kong S.K., and Hwang R. N. (1999). “Protecting adjacent DOI: 10.1061/(ASCE)1090-0241(2005)131:2(162).
buildings during underground construction.” Proceedings of Symposium South Branch of Institute for Building Science and Technology (IBST/
on Innovative Solutions in Structural and Geotechnical. S). Geotechnical monitoring report (2008). Saigon Pearl Apartment
Moormann, C. (2004). “Analysis of wall and ground movements due to buildings; (2009): Petro land tower, Blooming Park building;
deep excavations in soft soil based on a new worldwide database.” (2010): Sai Gon Times Square Project, Thanh Da View building,
Soils and Foundations, Japanese Geotechnical Society, Vol. 44, No. Sunrise City Plot V apartment buildings, Saigon MC towers, Le
1, pp. 87-98. Meridien hotel, Internal publications.
Nguyen (2010). The lesson from the collapse of the building of Institute Surarak, C., Likilersuang, S., Wanatowski, D., Balasubramaniam, A.,
of Social Sciences, Southern Region in Ho Chi Minh city. IBST Oh, E., and Guan, H. (2012). “Stiffness and strength parameters for
Technology Magazine, No. 3/2010, pp. 49-57. hardening soil model and stiff Bangkok clays.” Soil and Foundation,
Nguyen Minh CC&I (2012). Monitoring data of excavation – Vietcom Vol. 52, No. 4, pp. 682-697, DOI: 10.1016/j.sandf.2012.07.009.
bank tower, Vol. 1, 2, 3, 4, Internal publication Tan, Y. C. and Chow, C. M. (2008). Design of retaining wall and
Ou, C. Y. (2006). Deep Excavation: Theory and Practice, Taylor & support systems for deep basement construction – A Malaysian
Francis Group, London, UK, pp. 193-195. experience, Seminar on Deep Excavation and Retaining Walls,
Ou, C.Y., Hsieh, P. G., and Chiou, D. C. (1993). “Characteristics of Jointly organized by IEM-HKIE at Tropicana Golf & Country
ground surface settlement during excavation.” J. Can. Geotech, Vol. Resort, Petaling Jaya, Malaysia, pp. 12-14.
30, pp. 758-767. Terzaghi, K. (1943). Theoretical Soil Mechanics, John Wiley and Sons,
Pacific Consultants International in association with Oriental Consultants New York, N.Y.
Co., Ltd, Transportation Engineering Design Incorporation (TEDI), Twine, D. and Roscoe H. (1999). Temporary propping of deep
Environment Technology Co., Ltd. (2002). Saigon East West excavations – Guidance on design (C517D), CIRIA, London.
Highway project – Package 2C – Report on Soil Investigation. Ulrich, E. J. (1989). “Tieback supported cuts in overconsolidated soils.”
Pham, T. V. N (2008). Study on Land Subsidence due to Groundwater J. Geotech. Engrg., Vol. 115, No.4, pp. 521-545, DOI: 10.1061/
Exploitation in Ho Chi Minh city, Master’s Thesis. Asian Institute of (ASCE)0733-9410(1991)117:5(832.2).
Technology, Bangkok, Thailand, pp. 54-81. Yong, K. Y. and Lee, S. L. (2007). “Collapse of Nicoll Highway – A
Phienwej, N. and Gan, C. H. (2003). “Characteristics of ground movement in global failure at the Curved section of a cut-and-cover tunnel
deep excavations with concrete diaphragm walls in Bangkok soils construction (Chin Fung Kee lecture).” Proc. 16th SEAGC. Subang
and their prediction.” Journal of the Southeast Asian Geotechnical Jaya, Malaysia.

− 14 − KSCE Journal of Civil Engineering