開挖工程之回顧

陳嘉正
奧雅納工程顧問

OBSERVATIONS FROM EXCAVATIONS – A REFLECTION

Andrew K C Chan
Ove Arup & Partners Hong Kong Limited

撮要
本文介紹了兩組分別在香港及亞洲其他軟土地區進行的不同深度之基坑開
挖工程實 ，討論了在工程進行時所觀察到的牆身位移和影響其因素以及
預測未 開挖工程所需之設計参數等。
開挖工程所需之設計参數等 。

由於結果 自現場觀察，
自現場觀察，故較適用於與上述 子類似之開挖工程。
子類似之開挖工程。尤為重
要的是，
要的是 ，從累積的觀察結果中增加對設計方案之
從累積的 觀察結果中增加對設計方案之 心，
心，從而啟發創新及具
經濟效益之設計。
經濟效益之設計 。

最後引出兩 ，一 在香港的填海區，
在香港的填海區，另一 則在曼谷之軟土區，
則在曼谷之軟土區，作為本
文總結。
文總結。
 OBSERVATIONS FROM EXCAVATIONS – A REFLECTION
Andrew K C Chan 1

Abstract: Observed wall movements from two sets of case histories for a range of deep
excavations are presented. The first from excavations in Hong Kong and the second
from excavations on soft clay sites around Asia. These cases are discussed and general
observations made regarding factors affecting the observed movements and design
parameters relevant to the prediction of behaviour in future excavations. Observations
alone are only sufficient for future design if these designs are within the range of
experience represented by past cases. More importantly observations are needed to gain
confidence in design procedures so that these procedures can then be used to extrapolate
to innovative and more economical designs. The paper ends by showing two examples
where this has been done, one in a recent reclamation in Hong Kong and the other a soft
clay site in Bangkok.

INTRODUCTION
The geotechnical engineering design process can never rely on calculations alone.
Observations from previous similar projects in similar ground conditions enable these design
processes to be refined and unnecessary conservatisms to be removed. In this way confidence
in the design process can be developed leading to confidence in extrapolating the design
experience into more 'adventurous' and economic construction.

This paper presents a series of case histories that are associated with deep excavations.
The main emphasis observation presented here is the observed maximum lateral movement of
the retaining wall element during the excavation process. A series of observations contrasts
two types of ground profiles common in Southeast Asia. The first is a residual soil profile
such as found in Hong Kong where the soils have been weathered from the parent rock. As
such the stress history and the state of stress in the ground is difficult to quantify. In this case
a common design approach has been used and the selection of suitable soil parameters,
especially the soil stiffness empirically correlated with SPT 'N' values, is investigated by back
analysis of field measurements. The second is an alluvial or marine clay profile common in
the coastal areas of many Southeast Asian cities including Singapore, Bangkok, Taipei,
Shanghai and parts of Hong Kong. For these profiles, the deposition and stress history can be
assessed with some degree of accuracy and reasonable assessments made of the predicted
behaviour under changes in stresses due to engineering construction. Key factors that
influence the observed movements including the effectiveness of various construction
methods are discussed. The final part of the paper shows examples of where the design
process has been used to extrapolate previous experience to enable innovative and more
economical construction methods to be employed.

It must be stated that this paper is not intended as an exhaustive or even extensive
review of relevant projects. The author has deliberately selected projects that the author’s
company has been associated with over the last 20 years or so, and these have been selected to
illustrate the main principles of the use of case histories for the design of retaining walls for
deep excavations.
_______________________________
1
Chairman, Ove Arup & Partners Hong Kong Limited
DEEP EXCAVATIONS IN HONG KONG
The case histories of deep excavations in Hong Kong cover a wide variety of ground
conditions. Most involve a thickness of fill that overlies residual soil, usually completely
decomposed granite. In a couple of cases there is a layer of marine or alluvial deposits over
the residual soil. The fill varies significantly from a thin superficial layer of disturbed ground
to 20m thick layer of engineered fill placed as part of a reclamation process. The case
histories are presented in order of their construction and are as follows.

Chater Station, Central

The paper by Davies and Henkel (1980) describes this project, now referred to as
‘Central Station’, in some detail. Many geotechnical difficulties were encountered during the
construction. Significant settlement was observed to adjacent structures primarily due to the
effects of dewatering but was also due to the installation of the diaphragm walls panels and
the subsequent excavation. As this paper is concentrating on the effects of excavation only
the lateral wall movements associated with the excavation are presented.

Figure 1 shows a summary of the ground conditions that comprise 8m of fill over 8m
of marine deposits over completely decomposed granite (CDG). The excavation at this site
was 26m deep and a 1.2m thick diaphragm wall was used to support the excavation. The
station was excavated using the top-down method with the permanent station floor slabs
supporting the retaining walls as the excavation progressed downwards. The observed wall
movement is also shown in the figure. Unfortunately the movements were only recorded after
the upper slab of the station, the roof, was constructed.

The wall has been re-analyzed using the Oasys program FREW (Pappin et al 1986) to
determine the soil stiffness values that give rise to best agreement to those observed. This
computer program represents the wall as line of linear elastic bending beam finite elements
and attaches the wall to a block of soil on either side of the wall (see Figure 2). The stiffness
of the soil block is assessed using a method based on two dimensional finite element analyses.
The soil stresses are limited to remain within the active and passive force limits and internal
arching within the soil mass is allowed. Groundwater pressures are also modeled in the
analysis.

In common with standard practice in Hong Kong the soil Young’s Modulus E value
has been expressed as a function of the SPT N value (blows per 300mm). As shown in Figure
1 good agreement with observed measurements for the remaining stages of the excavation are
observed when an E value of 1.5N(MPa) is used for the Fill and Marine deposits, E=2N(MPa)
for the CDG, and E=4N(MPa) for CDG/HDG.
 Wall: 1.2m thick diaphragm wall
Deflection (mm) SPT N value
0 20 40 60 80 100 0 50 100 150 200
0 0

Fill E = 1.5N

-10 -10
Marine E = 1.5N

V.Weathered CDG E = 2N
-20 -20
Depth (m)

Dep th ( m)
-30 -30
CDG/HDG E = 4N

-40 -40

MDG
-50 -50
Measured
Calculated
Props
-60 -60

Figure 1 - Chater Station

Figure 1 – Chater Station

Strut

Figure 2 – Wall and Soil Model Used in Oasys FREW

DEEP EXCAVATIONS IN SOFT CLAY IN ASIA
Sites comprising a significant thickness of soft silty clay are common in the coastal
cities of Southeast Asia. Many of these alluvial or marine deposits have similar
characteristics. The case histories are from Singapore, Taipei and Shanghai and are again
presented in order of construction. The thickness of the soft clay layer is the greatest in
Shanghai.

Newton Station, Singapore

This project has previously been published by Nicholson (1987) and Gaba (1990).
The ground conditions are illustrated in Figure 10. The overall depth of excavation was
14.5m and a 0.8m thick diaphragm wall was used to support the retained soil. The method of
excavation was a top-down construction with permanent concrete floor slabs that also acted as
props to the wall, being cast as the excavation progressed. The excavation was initiated at the
eastern end of the station and a typical observed lateral movement of the retaining wall is
shown in Figure 10. The maximum wall movement is about 0.7% of the excavation depth.

Near the western end of the station, the marine clay became significantly thicker as
shown in Figure 11. An added complication was the presence of neighboring buildings
becoming quite close to the station. The observed movements at the eastern end, while being
somewhat less than those predicted would not be tolerable at the western end and the presence
of deeper clay would have meant that unacceptably large building displacements would be
likely to occur. A jet grout raft, installed prior to commencing the excavation, was therefore
adopted in that area. Figure 11 shows the depth of the jet grout raft and the movement that
resulted from the excavation at that area. It can be seen that the jet grout raft was very
effective in reducing the movement. The wall movement in this case is about 0.2% of the
excavation depth.

Deflection (mm) Deflection (mm)

Depth (m) Depth (m)
+102.6 0 20 40 60 80 100 120 +103.0 0 20 40 60 80 100 120
0 0
Fill Fill

-5 Upper Marine -5
Clay Marine
Clay
-10 -10

+88.1 +88.0
Clays
-15 -15
Fluvial Lower Marine Clay
0.8m Jet Grout 3.5m
Sand Compl etely
-20 -20 Decomposed
Completely Decomposed Granite (G4 ) 0.8m
Granite
-25 -25 Completely
Decomposed Granite
Case A I3
(G4)
Design Key: -30
-30 N N
Key:
Predicted By
Entrance Inclinometer
-35 Using Case A -35 W12
(With Jet
Newton Station Inclinometer I3 Newton Station
Grout )
W12 Scott R oad
(28-06-85) Scott R oa d
I3 W12 I3

Slabs Slabs
Entrance Entranc e

Figure 10 – Predicted and Measured Wall Figure 11 - Measured Wall Deflection

Deflection without Jet Grout (Newton Station) with Jet Grout (Newton Station)
Station BL14 of Nankang Line, Taipei
Several stations (BL13 to 16) of the Nankang Line were designed between 1989 and
1991 and constructed in the early to mid-90's. These involved deep excavations in a thick bed
of silty clay and other alluvial materials. The project area lies in the so-called K1 Zone of the
Keelung deposits and the geotechnical characteristics of the clay layers known as the
Sungshan Formations are described by Woo and Moh (1990).

Figure 12 shows a section of station BL14 and the foundations of adjacent buildings.
The construction required an excavation of 17.7m, supported by 1.2m thick diaphragm walls
and essentially by top-down method. Stringent limits on permissible movements were
imposed (e.g. 25mm) by the Client. At the start of the design process, three local case
histories on projects in the same area were back-analyzed using Oasys FREW to calibrate
parameters to be used. Early analysis, and indeed from empirical rule, e.g. Clough et al
(1989), indicates maximum wall deflection in the order of 80 to 90 mm and this when
translated into building settlements would be unacceptable. Having considered various
methods to reduce ground movements, and from the experience of Newton Station in
Singapore, it was decided to introduce a 3m thick jet-grout raft at below the excavation level.

Figure 13 shows the simplified soil engineering profile and the computed lateral wall
deflection profiles with and without the jet-grout raft. The actual "undrained" deflection
profile measured by a typical inclinometer is also shown. The jet-grout raft was effective in
reducing the lateral wall deflection to about 22mm and contributed significantly to the
protection of the sensitive buildings. The value of max/H measured is about 0.12%, being
well below the normal range measured from other projects in Taipei with no jet grouting.
Further data from this and other stations indicated rational analysis by FREW gave close
approximation of measured wall deflection profiles.

4 Storey Building 7 Storey Building

mPD mPD
+110 +110

+100 +100
Sungshan Formation
Upper Zone
Elevation Metres RL

3m

Jet Grout +90

+90
+85.0

Sungshan Formation
1.2m 1.2m Lower Zone +80
+80
+75.0
Chingmei
+73.0
Formation
+70 +70
Bedrock

Figure 12 – Section
Station of Station BL14
Section
 Undrained Shear Strength Deflection (mm)
Cu (kPa)
+108.6 0 20 40 60 80 0 10 20 30 40 50 60 70 80 90 100
+110

+105
Best Fit Field
Vane Shear
+100 Tests

+95 CL
Cu = 0.3Po’ +91.1
Elevation (m)

+90
Jet Grout 3m
1.2m
+85
Best Fit
+80 UU & UC
CL/SM Tests

+75

GM
+70

+65 Key:

+60 Predicted (without Jet Grout)

LEGEND: Best Fit UU &
CU Tests Predicted (with Jet Grout)
Cu = 0.3Po’
Inclinometer SID5 (with Jet Grout)
Best Fit Field Vane
Shear Tests Slabs

Struts

Figure 13 - Predicted and Measured Wall Deflection (Station BL14, Taipei)

New World Centre, Shanghai

A very large number of building projects were being constructed in the mid to late
1990's in Shanghai. The 250m tall New World Centre, located at the heart of the Lu Wan
District, is a typical example. General ground conditions have been described by Chan et al
(1996), and this project involved excavation of a 3-level basement. The site is along the busy
Huai Hai Road Central, and the Shanghai MRT Line 1 Huang Pi Station and the running
tunnels are at a distance of about 10m from the site boundary (see Figure 14).

Early work on the Shanghai Metro (Liu, 1993) and building basement excavations (e.g.
Mak and Liu, 1996) in the soft Shanghai clays indicate that large ground movements can
occur particularly where workmanship was less than ideal. In the case of the New World
Centre basement, it was decided to carry out the excavation in two phases (Figure 14). Phase
I, further away from the station and tunnels to avoid potentially excessive movements
affecting the metro facilities, has an excavation depth of 17.7m, was constructed first to allow
the tower erection to proceed. To reduce ground movements, it was further decided to
introduce a buried 'waling' of 6m thickness using deep mixed cement columns as shown in the
figure.

Figure 15 shows results of FREW analyses and measured wall deflection profiles.
The maximum measured deflection was about 40mm. When Phase I was complete, Phase II
excavation (for building podium area), two meters shallower at 15.7m, started with a jet-
grouted layer (buried 'prop') below excavation level covering a large part of the site. This
stronger measure was taken to further ensure effects on the metro works are minimized.
Figure 15 shows the jet grouting was effective in limiting the maximum deflection to about
20mm and the project was successfully completed without causing any undesirable problem.
The value of max/H with ground treatment measured at various parts of the site is very low
at about 0.12% to 0.23%.

Tunnel 700mm Dia Deep Mixed

Cement Piles @ 500mm c/c
Huang Pi Road South 500 500 500 500 500 500 (w/c = 0.45 & Cement
Content 130 kg/m3)

700
10m

Huang Pi Statio n
X7 Deep Mixed Cement Piles

tral
Jin Ling Road West

1000mm Dia Jet

Huai Hai Road Cen

Column @ 600mm c/c
10m 10m (UCS not less than

Phase I Phase II 600 600 600 600 600 600

1.2MPa f or Treated
Soil)
X12 20m
X18
A

1000
A
Deep Mixed Cement Piles Jet Grout

Jet Grout
10m Ground Treatment Details
Key:
Ma Dang Road
Inclinometer
Deep Mixed Cement Piles
Site Plan
Jet Grout
Phase I Phase II
0
5
10
Depth (m)

15 -15.73
-17.73
20 6m 8m Jet Grout
6m
25 Huang Pi Station
10m Deep Mixed Cement Piles 10m
30
35 0 10 20 30 40 50
40
Section A-A METRES

Figure 14 – Shanghai New World Centre

Undrained Shear Strength Deflection (mm) Deflection (mm)

Cu (kPa)
0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 60 70 80
0 20 40 60 80
0 Fill 1
Silt y Clay 2
Best Fit
-5 Mucky Silty Clay 3 Field Vane
Shear
Tests

Mucky Clay 4
-10

-15.73
-15
Depth (m)

-17.73
Clay 5 1
6m 8m
-20 Jet Grout
Silty Clay 5 2

-25 Cu = 0.3Po’ 10m

1m

-30 1m
Silty Clay With Deep Mixed Cement
Sand 5 3 Piles
-35

Phase I Excavation Phase II Excavation

-40 Key: Key:
LEGEND:
Inclinometer X7
(With Deep Mixed Cement Piles)
Inclinometer X18
C u (UU Test)
Inclinometer X 12 (WithJet
(With Deep Mixed Cement Piles)
Grout)
C u = 0.3Po’ (With Deep Mixed Cement Piles)
Predicted
Best Fit Fiel d Predicted
Vane S hear (Without Deep Mixed Cement Piles) (With Jet Grout)
Tests
Predicted Predicted
(With Deep Mixed Cement Piles) (Without Jet Grout)

Struts Struts

Figure 15 – Predicted and Measured Wall Deflection

Discussion
Excavations carried out in soft silty clay deposits can cause relatively large wall
movements. Experience from different sites in Asian cities shows similarity and consistency
in observed behaviour. Empirical rules such as Clough and O'Rourke (1990) give reasonable
first estimate of maximum wall movement. Analyses using FREW with a set of soil
parameters from standard measurements (e.g. strength) and from empirical correlations (e.g.
stiffness, earth pressure coefficients) can give good enough predictions of the maximum
movement and profile. Clearly, more high quality monitoring data specific to each location
will enhance the confidence in our ability to predict, perhaps with more sophisticated
numerical analysis and soil stress-strain models.

It should be noted that movements in soft clay profiles are particularly affected by
actual details of the excavation method and workmanship, and may be time-dependent. To
be of real value, case histories need to record as much relevant details as possible.

Stringent requirements are often set by building control or mass transit authorities for
protection of properties adjacent to deep excavations, some being absolutely necessary. In
clay profiles, these movement limits are difficult if not impossible to satisfy and ground
treatment, such as jet grouting before excavation, becomes necessary. From the cases
presented, it is shown that such strengthening or 'buried props' can be effective in achieving a
max/H of about 0.2% or below.

EXTRAPOLATION TO ENABLE INNOVATIVE DESIGN

In this section two examples are presented where established design methods,
calibrated against previous experience, have been used to produce innovative and economical
solutions in problems involving deep excavations. The first is in a recent reclamation in Hong
Kong, followed by one in Bangkok.

Tseung Kwan O Station, Hong Kong

This project, previously published by Ho et al (2001), involved the use of precast
concrete props in deep trenches such that they were in position just below the final excavation
level. The ground profile comprised about 12m thickness of recent reclamation fill over 10m
of soft marine clay over 20m of fine and coarse alluvium over a varying thickness of
completely decomposed volcanic residual soil. Originally the 1.2m thick diaphragm wall was
designed by Arup to use a jet grout raft just below the deepest excavation. The jet grout raft
was required to limit the bending moments in the wall to manageable levels rather than to
restrict lateral deformations.

The Contractor, Leighton-China State JV, proposed an alternative scheme comprising

underwater excavation. The wall was constructed and the excavation commenced. At this
stage it became apparent that the marine clay, which originally had not reached full primary
consolidation under the action of the recent fill, was steadily gaining strength as it approaches
normal consolidation. The contractor wished to explore the possibility of eliminating the
underwater excavation scheme and approached Arup to develop a more conventional
alternative scheme. The use of precast concrete props lowered into 7m deep trenches, placed
just below the final excavation level and secured to the wall with tremie concrete was jointly
developed between Arup and the contractor. Figure 16 shows a schematic construction
sequence for this work. Again this sequence of temporary propping was required to limit the
bending stresses within the already completed diaphragm wall.
DEEP EXCAVATIONS IN SOFT CLAY IN ASIA
Sites comprising a significant thickness of soft silty clay are common in the coastal
cities of Southeast Asia. Many of these alluvial or marine deposits have similar
characteristics. The case histories are from Singapore, Taipei and Shanghai and are again
presented in order of construction. The thickness of the soft clay layer is the greatest in
Shanghai.

Newton Station, Singapore

This project has previously been published by Nicholson (1987) and Gaba (1990).
The ground conditions are illustrated in Figure 10. The overall depth of excavation was
14.5m and a 0.8m thick diaphragm wall was used to support the retained soil. The method of
excavation was a top-down construction with permanent concrete floor slabs that also acted as
props to the wall, being cast as the excavation progressed. The excavation was initiated at the
eastern end of the station and a typical observed lateral movement of the retaining wall is
shown in Figure 10. The maximum wall movement is about 0.7% of the excavation depth.

Near the western end of the station, the marine clay became significantly thicker as
shown in Figure 11. An added complication was the presence of neighboring buildings
becoming quite close to the station. The observed movements at the eastern end, while being
somewhat less than those predicted would not be tolerable at the western end and the presence
of deeper clay would have meant that unacceptably large building displacements would be
likely to occur. A jet grout raft, installed prior to commencing the excavation, was therefore
adopted in that area. Figure 11 shows the depth of the jet grout raft and the movement that
resulted from the excavation at that area. It can be seen that the jet grout raft was very
effective in reducing the movement. The wall movement in this case is about 0.2% of the
excavation depth.

Deflection (mm) Deflection (mm)

Depth (m) Depth (m)
+102.6 0 20 40 60 80 100 120 +103.0 0 20 40 60 80 100 120
0 0
Fill Fill

-5 Upper Marine -5
Clay Marine
Clay
-10 -10

+88.1 +88.0
Clays
-15 -15
Fluvial Lower Marine Clay
0.8m Jet Grout 3.5m
Sand Compl etely
-20 -20 Decomposed
Completely Decomposed Granite (G4 ) 0.8m
Granite
-25 -25 Completely
Decomposed Granite
Case A I3
(G4)
Design Key: -30
-30 N N
Key:
Predicted By
Entrance Inclinometer
-35 Using Case A -35 W12
(With Jet
Newton Station Inclinometer I3 Newton Station
Grout )
W12 Scott R oad
(28-06-85) Scott R oa d
I3 W12 I3

Slabs Slabs
Entrance Entranc e

Figure 10 – Predicted and Measured Wall Figure 11 - Measured Wall Deflection

Deflection without Jet Grout (Newton Station) with Jet Grout (Newton Station)
MRTA Northern Section, Bangkok
An innovative scheme was adopted in this project to optimize the construction process
of this major metro construction project in Bangkok, see Davies et al (2001). Tunnel Boring
Machines were driven through three station boxes prior to the completion of the excavation
within these boxes. Figure 18 shows the work arrangement at this stage. To facilitate
unimpeded construction progress, tunnel “eyes”, reinforced with glass fibre reinforcement
polymer, were designed at the end walls of the station box. The construction clearance
between the main diaphragm walls and the tunnel and between the internal stanchions and the
tunnel was about 1m. After the tunnels were completed the station excavation was progressed
using the top down method and eventually the tunnels within the station box demolished.
Based on experience of using FREW, a detailed finite element numerical analysis was carried
out to evaluate the interaction between the tunnel, the walls and the station stanchions and the
effects of excavation within the station box on these elements. Figure 19 shows the analysis
of the tunnelling stage of the excavation. The site investigations carried out for the design
included self boring pressure meter testing to determine the small strain stiffness of the soil
profile and give appropriate values for the station design. The field and laboratory test
results, in combination with the numerical analysis allowed the design to be expeditiously
undertaken and the tunnelling to proceed with confidence. This is the first time concurrent
tunnelling and station construction has been undertaken on this scale.

Station work area

Temporary hording

Temporary deck Roof level

Bangkok Soft Clay

Depth of
excavation
Bangkok Stiff Clay

Bangkok Aquifer

Figure 18 – Construction Arrangement of Bangkok MRTA

 Figure 19 – Numerical Analysis Result at Tunnelling Stage

CONCLUSIONS
The above case histories of showing observed maximum lateral movements of
retaining walls due to deep excavations forms a valuable resource of previous experience in
these problems. They show that a combination of good quality field data and rational analysis
can give the engineer confidence in the design predictions.

They also lead to the following observations:

1. Well-controlled deep excavations in urban Hong Kong are likely to cause maximum
lateral displacements of the retaining walls that will be between 0.15% to 0.3% of the
final excavation depth.

2. Back analysis of excavation lateral movements in Hong Kong shows that for
movement prediction purposes, reasonable and conservative estimates of displacement
are likely to be achieved by using an Young’s Modulus (E) for soil of 1.5N(MPa) for
fill and alluvium and a value of E = 2N(MPa) for completely decomposed granite.
Higher E/N values are relevant for CDG/HDG with high 'N' values.
3. Experience from soft clay sites is widely applicable and can be used with reasonable
confidence for sites around the region.

4. Observations in themselves are only sufficient for a robust design if the current design
situation fits within the range of previous experience. To enable extrapolation to
larger and/or more economical design, it is necessary to use back analysis to validate
and give confidence in the employed design methods. This process enables these
design methods to be used to extrapolate to innovative and economical designs with
some confidence.

REFERENCES
Chan A.K.C., Lui, J.Y.H., Yau, K.F. and Yin, K.K. (1996) "Aspects of Geotechnical Practice
in China with particular reference to foundations in Shanghai" Proc. of the 16th HKIE
Geotechnical Division Annual Seminar.
Clough, R.W. and O’Rourke, T. D. (1990). “Construction induced movements of insitu walls”.
Proc. ASCE Conf. on Design and Performance of Earth Retaining Structures, Geotech.
Spec. Publ. No. 25, pp. 439-470.
Clough, R.W., Smith, E.M. and Sweeney, B.P. (1989) “Movement control of excavation
support systems by iterative design”, Proc. ASCE Foundation Engineering: Current
Principles and Practices., Vol. 2, ASCE, New York, pp 869-884.
Davies, J.A. (1987) "Groundwater control in the design and excavation of a deep excavation".
Proc. of the 9th European Conference on Soil Mechanics & Foundations Engineering,
Dublin, pp139-144.
Davies, J., Thompson, P. and Young, S.T. (2001) “A comparison between tender, detailed
design and the field performance of diaphragm walls in Bangkok”. Proc. of the 14th
Southeast Asian Geotechnical Conference.
Davies, R.V. and Henkel, D.J. (1980). “Geotechnical Problems Associated with Construction
of Chater Station”. Proc. of the Conference on Mass Transportation in Asia, Hong
Kong.
Gaba, A.R. (1990). “Jet Grouting at Newton Station, Singapore”. Proc. 10th Southeast Asian
Geotechnical Conf. Vol. 2, pp77-79.
Ho J., Hope S., Pappin J.W. and Blair C (2001). “Buried Concrete Prop for Tseung Kwan O
Station and Tunnels”. Proc. of the 20th HKIE Geotechnical Division Annual Seminar.
Humpheson, C., Fitzpatrick, A.J.& Anderson J.M.D. (1986). “The basements and substructure
for the new headquarters of the Hongkong and Shanghai Banking Corporation, Hong
Kong”. Proc. Institution of Civil Engineers, Vol. 80, London, pp851-883.
Lee, D.M., Pappin, J.W. and Buckler, J.G. (2001). “Control of ground movement in the
construction of Festival Walk”. Proc 14th Southeast Asian Geotechnical Conference.
Liu, J.H. (1993) “The Prediction of Ground Movements around Slurry Wall Supported Deep
Excavations”. Proc. Sino-American Technology Engineering Conference,
Infrastructure Construction Session, Beijing, pp84-85.
Long, M. (2001). “Database for Retaining Wall and Ground Movements due to Deep
Excavations”. ASCE Journal of Geotechnical and GeoEnvironmental Engineering, pp.
203-224.
Lui, J.Y.H. and Yau, P.K.F. (1995). “The Performance of the Deep Basement for Dragon
Centre”. Proc. of the 15th HKIE Geotechnical Seminar on Instrumentation in
Geotechnical Engineering.
Mak, L.M. and Liu, J.H. (1996) “Geotechnical problems associated with substructures
construction above Metro Tunnels” Proc. of the 16th HKIE Geotechnical Division
Annual Seminar.
Nicholson, D.P. (1987) “ The Design and Performance of the retaining Wall at Newton
Station”. Proc. of the Singapore MRTC Conference, Singapore, pp147-154.
Pan J.K.L., Pappin J.W., Cowan S. & Lam L. W. Y. (2001) “An Application of the
Observational Method at Tseung Kwan O Station and Tunnels”. Proc. of the 20th HKIE
Geotechnical Division Annual Seminar.
Pappin, J.W., Simpson, B., Felton, P.J. and Raison, C. (1986). “Numerical analysis of
flexible retaining walls” Symposium on computer applications in geotechnical
engineering, The Midland Geotechnical Society, April.
Sze, W.C.J. and Young, S.T. (2003). “Design and Construction of a Deep Basement through
an existing Basement at Central”, Proc. of the 22nd HKIE Geotechnical Division
Annual Seminar.
Walsh, N.M. and Fung, K. (1989) “The Use of Inclinometers to Improve the Modeling of
Retaining Walls”. Proc of the 9th HKIE Geotechnical Division Annual Seminar.
Woo, S.M. and Moh, Z.C. (1990) "Geotechnical Characteristics of Soils in the Taipei Basin",
Proc. of the 10th Southeast Asian Geotechnical Conference, Taipei, Volume II.
Young, S.T. (2003). “Slurry Wall Design for Hong Kong Station”. Proc. of Conference
“Earth Retention Systems 2003” jointly organised by ASCE Metropolitan Section
Geotechnical Group, The Deep Foundation Institute and ADSC: The International
Association of Foundation Drilling, May 2003, New York City.

ACKNOWLEDGEMENT
The measurements for the cases in Taipei and Shanghai are taken from internal files of
Arup Hong Kong. Assistance from many colleagues in Arup Geotechnics, particularly Dr
Jack Pappin and Ir Mark Choi, is gratefully acknowledged.
 The scheme progressed using the precast beams. The marine clay continued to gain
strength and the observed movements of the wall showed that the predicted displacements
were too large. A back analysis procedure using FREW was therefore adopted and revised
soil stiffness parameters developed to match the observed displacement. These revised
parameters, described in details in Pan et al (2001), and illustrated in Figure 17, were then
used in an application of the Observational Method to eliminate the buried prop in the latter
part of the excavation. All of this work was carried out under the Buildings Department
approval procedure.
Temporary Struts

Diaphragm Wall Diaphragm Wall

Tunnel Box
Structure
Slurry
Buried Trenches
Concrete Buried
Prop Concrete
Bored Piles Prop

Figure 16 - Excavation sequence at Tseung Kwan O Station

Tseung Kwan O Station Wall: 1.2m thick diaphragm wall

Deflection (mm) SPT N value
0 20 40 60 80 100 0 50 100 150 200
0 0

Fill E = 1.2N
-10 -10

-20 -20 Marine

E = 500cu
clay
Depth (m)

Dep th ( m)

Fine
-30 -30 E = 500cu
Alluvium

Coarse Alluvium E = 4N
-40 -40

-50 -50 CDV E = 4N

Measured
Calculated
Props
MDV
-60 -60
0 50 100 150 200
cu (kPa)
shown dashed

Figure 17 - Tseung Kwan O Station

MRTA Northern Section, Bangkok
An innovative scheme was adopted in this project to optimize the construction process
of this major metro construction project in Bangkok, see Davies et al (2001). Tunnel Boring
Machines were driven through three station boxes prior to the completion of the excavation
within these boxes. Figure 18 shows the work arrangement at this stage. To facilitate
unimpeded construction progress, tunnel “eyes”, reinforced with glass fibre reinforcement
polymer, were designed at the end walls of the station box. The construction clearance
between the main diaphragm walls and the tunnel and between the internal stanchions and the
tunnel was about 1m. After the tunnels were completed the station excavation was progressed
using the top down method and eventually the tunnels within the station box demolished.
Based on experience of using FREW, a detailed finite element numerical analysis was carried
out to evaluate the interaction between the tunnel, the walls and the station stanchions and the
effects of excavation within the station box on these elements. Figure 19 shows the analysis
of the tunnelling stage of the excavation. The site investigations carried out for the design
included self boring pressure meter testing to determine the small strain stiffness of the soil
profile and give appropriate values for the station design. The field and laboratory test
results, in combination with the numerical analysis allowed the design to be expeditiously
undertaken and the tunnelling to proceed with confidence. This is the first time concurrent
tunnelling and station construction has been undertaken on this scale.

Station work area

Temporary hording

Temporary deck Roof level

Bangkok Soft Clay

Depth of
excavation
Bangkok Stiff Clay

Bangkok Aquifer

Figure 18 – Construction Arrangement of Bangkok MRTA

 Figure 19 – Numerical Analysis Result at Tunnelling Stage

CONCLUSIONS
The above case histories of showing observed maximum lateral movements of
retaining walls due to deep excavations forms a valuable resource of previous experience in
these problems. They show that a combination of good quality field data and rational analysis
can give the engineer confidence in the design predictions.

They also lead to the following observations:

1. Well-controlled deep excavations in urban Hong Kong are likely to cause maximum
lateral displacements of the retaining walls that will be between 0.15% to 0.3% of the
final excavation depth.

2. Back analysis of excavation lateral movements in Hong Kong shows that for
movement prediction purposes, reasonable and conservative estimates of displacement
are likely to be achieved by using an Young’s Modulus (E) for soil of 1.5N(MPa) for
fill and alluvium and a value of E = 2N(MPa) for completely decomposed granite.
Higher E/N values are relevant for CDG/HDG with high 'N' values.
3. Experience from soft clay sites is widely applicable and can be used with reasonable
confidence for sites around the region.

4. Observations in themselves are only sufficient for a robust design if the current design
situation fits within the range of previous experience. To enable extrapolation to
larger and/or more economical design, it is necessary to use back analysis to validate
and give confidence in the employed design methods. This process enables these
design methods to be used to extrapolate to innovative and economical designs with
some confidence.

REFERENCES
Chan A.K.C., Lui, J.Y.H., Yau, K.F. and Yin, K.K. (1996) "Aspects of Geotechnical Practice
in China with particular reference to foundations in Shanghai" Proc. of the 16th HKIE
Geotechnical Division Annual Seminar.
Clough, R.W. and O’Rourke, T. D. (1990). “Construction induced movements of insitu walls”.
Proc. ASCE Conf. on Design and Performance of Earth Retaining Structures, Geotech.
Spec. Publ. No. 25, pp. 439-470.
Clough, R.W., Smith, E.M. and Sweeney, B.P. (1989) “Movement control of excavation
support systems by iterative design”, Proc. ASCE Foundation Engineering: Current
Principles and Practices., Vol. 2, ASCE, New York, pp 869-884.
Davies, J.A. (1987) "Groundwater control in the design and excavation of a deep excavation".
Proc. of the 9th European Conference on Soil Mechanics & Foundations Engineering,
Dublin, pp139-144.
Davies, J., Thompson, P. and Young, S.T. (2001) “A comparison between tender, detailed
design and the field performance of diaphragm walls in Bangkok”. Proc. of the 14th
Southeast Asian Geotechnical Conference.
Davies, R.V. and Henkel, D.J. (1980). “Geotechnical Problems Associated with Construction
of Chater Station”. Proc. of the Conference on Mass Transportation in Asia, Hong
Kong.
Gaba, A.R. (1990). “Jet Grouting at Newton Station, Singapore”. Proc. 10th Southeast Asian
Geotechnical Conf. Vol. 2, pp77-79.
Ho J., Hope S., Pappin J.W. and Blair C (2001). “Buried Concrete Prop for Tseung Kwan O
Station and Tunnels”. Proc. of the 20th HKIE Geotechnical Division Annual Seminar.
Humpheson, C., Fitzpatrick, A.J.& Anderson J.M.D. (1986). “The basements and substructure
for the new headquarters of the Hongkong and Shanghai Banking Corporation, Hong
Kong”. Proc. Institution of Civil Engineers, Vol. 80, London, pp851-883.
Lee, D.M., Pappin, J.W. and Buckler, J.G. (2001). “Control of ground movement in the
construction of Festival Walk”. Proc 14th Southeast Asian Geotechnical Conference.
Liu, J.H. (1993) “The Prediction of Ground Movements around Slurry Wall Supported Deep
Excavations”. Proc. Sino-American Technology Engineering Conference,
Infrastructure Construction Session, Beijing, pp84-85.
Long, M. (2001). “Database for Retaining Wall and Ground Movements due to Deep
Excavations”. ASCE Journal of Geotechnical and GeoEnvironmental Engineering, pp.
203-224.
Lui, J.Y.H. and Yau, P.K.F. (1995). “The Performance of the Deep Basement for Dragon
Centre”. Proc. of the 15th HKIE Geotechnical Seminar on Instrumentation in
Geotechnical Engineering.
Mak, L.M. and Liu, J.H. (1996) “Geotechnical problems associated with substructures
construction above Metro Tunnels” Proc. of the 16th HKIE Geotechnical Division
Annual Seminar.
Nicholson, D.P. (1987) “ The Design and Performance of the retaining Wall at Newton
Station”. Proc. of the Singapore MRTC Conference, Singapore, pp147-154.
Pan J.K.L., Pappin J.W., Cowan S. & Lam L. W. Y. (2001) “An Application of the
Observational Method at Tseung Kwan O Station and Tunnels”. Proc. of the 20th HKIE
Geotechnical Division Annual Seminar.
Pappin, J.W., Simpson, B., Felton, P.J. and Raison, C. (1986). “Numerical analysis of
flexible retaining walls” Symposium on computer applications in geotechnical
engineering, The Midland Geotechnical Society, April.
Sze, W.C.J. and Young, S.T. (2003). “Design and Construction of a Deep Basement through
an existing Basement at Central”, Proc. of the 22nd HKIE Geotechnical Division
Annual Seminar.
Walsh, N.M. and Fung, K. (1989) “The Use of Inclinometers to Improve the Modeling of
Retaining Walls”. Proc of the 9th HKIE Geotechnical Division Annual Seminar.
Woo, S.M. and Moh, Z.C. (1990) "Geotechnical Characteristics of Soils in the Taipei Basin",
Proc. of the 10th Southeast Asian Geotechnical Conference, Taipei, Volume II.
Young, S.T. (2003). “Slurry Wall Design for Hong Kong Station”. Proc. of Conference
“Earth Retention Systems 2003” jointly organised by ASCE Metropolitan Section
Geotechnical Group, The Deep Foundation Institute and ADSC: The International
Association of Foundation Drilling, May 2003, New York City.

ACKNOWLEDGEMENT
The measurements for the cases in Taipei and Shanghai are taken from internal files of
Arup Hong Kong. Assistance from many colleagues in Arup Geotechnics, particularly Dr
Jack Pappin and Ir Mark Choi, is gratefully acknowledged.