Geotechnical Engineering

Volume 167 Issue GE2

Barangaroo South harbourside basement,
Australia, challenges and solutions
Badelow, Wakefield, Chandrasegaran and Iseppi

Proceedings of the Institution of Civil Engineers

Geotechnical Engineering 167 April 2014 Issue GE2
Pages 194204 http://dx.doi.org/10.1680/geng.13.00031
Paper 1300031
Received 26/03/2013
Accepted 12/09/2013
Published online 02/12/2013
Keywords: anchors & anchorages/diaphragm walls/excavation
ICE Publishing: All rights reserved

Barangaroo South harbourside

basement, Australia,
challenges and solutions
1
j
Frances Badelow BEng, CEng, MICE

3
j
Chandra Chandrasegaran MSc, PE (Geotechnical)

Senior Principal, Coffey Geotechnics Pty Ltd, Sydney, Australia

Singapore and Malaysia, PE (Civil) Singapore, MIEM

Design Manager, Menard Bachy Pty Ltd, Singapore

2
j
Malcolm Wakefield Ingenieur, ENSG

4
j
Glen Iseppi BE, CPEng, MIEAust

Design Engineer, Menard Bachy Pty Ltd, France

Senior Engineer, Brown Consulting Pty Ltd, Sydney, Australia

1
j

2
j

3
j

4
j

At Barangaroo South on Sydneys central business district waterfront, Australia, Lend Lease is working with the New
South Wales government on the A$6 billion (4 billion) transformation of a former container port into a thriving
business, residential and leisure precinct. The vision is that Barangaroo South will enhance Sydneys position as an
internationally appealing, globally competitive, environmentally sustainable city. The development will be Australias
first large-scale carbon dioxide neutral precinct. Barangaroo South will feature three commercial towers ranging in
height from 39 to 49 storeys, sharing a common, two-level basement to be retained around its perimeter by an
approximately 770 m long diaphragm wall, socketed into Sydney sandstone. As the basement excavation will be
below sea level, it is essential that the site retention wall for the basement also acts as a cut-off against groundwater
ingress. This paper provides an overview of the design and construction challenges facing the installation of the
diaphragm wall and the innovative solutions adopted to overcome these challenges.

1.

Introduction

Barangaroo South is a major mixed use development located at

the southern end of the Barangaroo development project in
Sydney, Australia (Wong et al., 2013). It is located on Sydneys
Central Business District waterfront on the foreshore of Darling

Harbour as shown in Figure 1, and has a site area of over 7.5 ha.
The vision of the developer for the project, Lend Lease, is that
Barangaroo South will enhance Sydneys position as an internationally appealing, globally competitive and environmentally
sustainable city. The development will be Australias first largescale carbon dioxide neutral precinct.
Located on the southern third of the 22 ha Barangaroo site, stage
1A of Barangaroo South will feature three commercial towers
ranging in height from 39 to 49 storeys, sharing a common, twolevel basement to be retained around its perimeter by an
approximately 770 m long diaphragm wall, socketed into Sydney
sandstone. As the basement excavation will be below sea level, it
is essential that the site retention wall for the basement also acts
as a cut-off against groundwater ingress. This had to be considered when designing the reinforcement and the temporary ground
anchor support required prior to construction of the permanent
floor slabs.

Figure 1. Location of Barangaroo South

194

A number of challenges had to be overcome during the design

and construction of the diaphragm wall, including difficulties
associated with old harbour walls and buried wharf structures,

Geotechnical Engineering
Volume 167 Issue GE2

Barangaroo South harbourside basement,

Australia, challenges and solutions
Badelow, Wakefield, Chandrasegaran and
Iseppi

steeply dipping rock levels and buried cliff lines, accommodating

the proposed future Sydney Metro tunnels, the close proximity of
existing building basements, as well as complex bulk excavation
staging. Stringent wall deflection and settlement criteria needed
to be met, as well as the need to consider seismic events and sea
level rise due to climate change.

redevelopment, Barangaroo was occupied by a shipping and

overseas passenger terminal, with the majority of the site covered
by hardstand, with areas of suspended concrete slab.

2.

Site history

Barangaroo has a long post-European settlement history of

industrial use, predominantly as wharfage but also as a gasworks.
The original, undeveloped natural shoreline comprised rocky
outcrops and mangrove swamps. The first wharves were built at
Millers Point in the mid-1820s and there are reports that parts of
Barangaroo were utilised as wharfage as early as 1788. The
gasworks were constructed in 1840 on the coastline of the time
and substantial reclamation of the area took place during the midand late nineteenth century, with a sandstone block harbour wall
and numerous wharves constructed at this time.
Following the demolition and backfilling of the gasworks in
1925, timber wharves were constructed in its place and further
northward, such that by the end of the 1930s the entire shoreline
in the site area was fronted by timber wharves extending from a
sandstone harbour wall. In the early 1970s a concrete caisson
harbour wall, which still forms the western boundary of the
current site, was constructed parallel to and some 150 m out from
the former harbour wall and the reclamation formed by end
dumping of fill. The timber finger wharf piers were left in place
and a variable depth of predominantly sandstone rubble fill up to
21 m was placed between the two harbour walls. Prior to

3.

Geological overview

The geological conditions at Barangaroo are complex. The site is

underlain by fill and Quaternary alluvium overlying Triassic age
Hawkesbury Sandstone a fine- to coarse-grained quartzose
sandstone deposited in 13 m thick beds. Structurally, there are two
main rock joint sets, one set trending NNE and an orthogonal set
trending ESE. The orientation of these joint sets and the bedding
spacing of the Hawkesbury Sandstone, together with erosion and
weathering processes in a foreshore environment, results in a steeply
stepping rock profile with overhangs, detached blocks and wide,
open sub-vertical joints around the harbour foreshore (Figure 2).
The presence of buried cliff lines results in a large variation in
rock level across the site, from about 1 m to 2 m below ground
level in the north-eastern corner of the site to over 30 m below
ground level in the north-western corner a distance of only
some 150 m. In addition, a NWSE trending palaeochannel
crosses the site, extending to a depth of approximately 18 m
below ground level and infilled with predominantly loose to
medium dense alluvial sand, interbedded with subordinate firm
clay. There is potential for the alluvial deposits to be acid sulfate
bearing, which may be a problem if they are allowed to oxidise
following exposure to air.
The fill material used for the land reclamation is highly variable,
both in nature and thickness; many buried obstructions, including
Detached slabby blocks

Overhang

Detached blocks

Wide, open joints

Weaker horizon
(e.g. shale/mudstone)
and associated undercutting

20 m (approx.)

Figure 2. Sandstone cliff profile

195

Geotechnical Engineering
Volume 167 Issue GE2

Barangaroo South harbourside basement,

Australia, challenges and solutions
Badelow, Wakefield, Chandrasegaran and
Iseppi

old and existing harbour walls and timber wharves, are present.
Large voids were also identified in the fill during the site
investigation works, as well as wood, steel, sandstone boulders
and other building rubble.

the loads to be carried by the ground anchors and floor slabs.

Each analysis typically included 13 stages in order to simulate
the temporary and long-term conditions. Check analyses using
the Menard Bachy computer software based on subgrade reaction
theory were completed to validate the Plaxis results.

A typical geological section along the western (harbourside)

boundary of the site is given in Figure 3. The deep fill profile
with thick underlying alluvial deposits will provide limited lateral
support for a deep foundation system. Also, the fill material
across the site is highly permeable and the near-surface water
table responds strongly to tidal variations. It is likely that
hydraulic connection is present between the Hawkesbury Sandstone and the harbour, although the rock mass permeability is
generally low to moderate (typically less than 2 uL) and governed
by defects in the rock.

As the development area is large, temporary internal propping of

the perimeter diaphragm wall was not considered to be effective,
except possibly in the corners of the site. A top-down construction solution was considered but not adopted as the client needed
to maintain flexibility in the excavation and construction sequencing, as the development layout had not been finalised at the time
of diaphragm wall design.

Owing to an extremely variable bedrock level and other site

constraints, design had to be carried out for a relatively large
number of wall cross-sections. Typical wall thicknesses were
0.8 m and 1.2 m with individual panel lengths varying from
3.4 m to 7.8 m, constructed using either 2.8 m wide excavation
grabs or Hydrofraise techniques, forming bites, with up to three
bites required to form a panel. The general excavation level is at
5.85 m Australian height datum (AHD) (approximately 8.5 m
below ground level), with some locally deeper excavations for
pile caps, pump rooms and lift pits. At tender stage, it was
decided to anchor the diaphragm wall with a single row of
prestressed multi-strand anchors, installed just above the groundwater table at +1.2 m AHD. This resulted in an increase in steel
reinforcement, as the anchors were positioned above the optimum
level for bending moment distribution but avoided the high-risk
operation of installing anchors below the groundwater table. The
basement general layout is shown in Figure 4 and typical
diaphragm wall section is shown in Figure 5.

The diaphragm wall was designed using the commercially available finite-element program Plaxis to assess structural actions and
deflections and associated adjacent ground movements, as well as

4.1 Design criteria

Stringent design criteria were specified for the diaphragm wall,
which included

4.

Design challenges and solutions

Excavation level RL 585 m AHD

BH214
Offset 75 m

10

?
?

HW/MW

N* 16
N* R

N* * 14
N 7
N* 6
N* 17
N* 24
N* 15
N* 0
N** 6
?N
5

BAR24
BAR14
Offset 107 m Offset 59 m

N* 14
N* 13
N* 12
N* 32
N* 11
N* R
N* * 11
? N R?

SW

?S?

FR

Bottom depth
1795 m

20

N* 18
N* 59
N* 31
N* 21
N* 32
N* R
?
N* 7
N* 19
N* 9
N* R

32 uL

?
MW
XW
FR
HW

BAR23
Offset 101 m
W8_BH 108A
Offset 68 m

N* 5
N* 1
N* 38
N* 21
N* R
N* 8
N* 33
N* 17
Sf
N* 6
L/WD ? *
? N * 25
N R
Bottom depth
? 1645 m?
?

FR

Bottom depth
2104 m

DW

BAR13
Offset 39 m

Fill
?
?

Bottom depth
212 m

?
?

20 uL

?F

Sf
S/F
D
SI

Alluvium
?
?

?
SW

FR

FR

N
N/M

Bottom depth
2555 m

Bottom depth
2635 m

No R
No 13
No 9
No 2
No R
No 11
No 9
No 5
No 14
No 53
No? 0
No 4
No 34
No 13
No 12
?
No R

?
?

HW
XV
VW
FR

Sf
D
MD

?
?

Bottom depth
2795 m

N* 15
N* 21
N* R
N* 9
N* 3
N* 18
N* 13
N* 43
N* 2
N* 14
*
F/Sf
?
? N * 22
N ?R
F
N* 23
MD
MD
N* 9
Sf
*
D N 20
MD
N* 18
VSf *
Sf/
N 26
*
o
2 MD N 27
? N MW
N* 15
SV
?
*
SW
MW/SW
N
24
FR
FR
N* R
Bottom depth Bottom depth MW
SW
2098
m
3164 m

N* 18
N* 12
N* 3
N* R
N* 11
N* 10
N* R
N* 25
N* 15
N* 12
*
?N* 22
N 10
N* 12
N* 12
N* 21
N* 24
N* 14
*
?N 20

30

HW
MW

N 22
N3

Weathered
Hawkesbury
Sandstone

N* 9
N* 2
N* 5
N* 13
N* 8
N* 9
N* 8
N*? 5
N* 9
N* 3
N* 0
N* 42
*
?N R

BAR17
Offset 46 m
BAR21
BAR12
Offset 122 m
Offset 49 m

W8_BH 103
BAR22
Offset 66 m Offset 127 m

300 uL

FR

?
MD
L/WD
MD
Sf/VSf
MD
VSf
MD

FR

SW
FR

Bottom depth
3882 m

40

Limit of site

Limit of site

50

Hawkesbury Sandstone

HW
FR

Bottom depth
55 m

60
0

50

Figure 3. Typical geotechnical section

196

100

150

200

250

Geotechnical Engineering
Volume 167 Issue GE2

Barangaroo South harbourside basement,

Australia, challenges and solutions
Badelow, Wakefield, Chandrasegaran and
Iseppi

P116

P2
6
P2
7
P2
8
P2
9
P3
0
P3
1

P2
5

P2
0
P2
1
P2
2
P2
3
P2
4

P1
8
P1
9

P1
7

P1
3
P1
4
P1
5
P1
6

800 mm temporary panels

with shear pins

P117

P1
1
P1
2

P1
0

P8
P9

P7

P6

P5

P4

P3

P1
P2

Future
development
foundations
P32
P33

800 mm panels with shear pins

P34

Trouserleg panels

Bored piles

P35
P36

P115

P3
7
P3
8
P3
9
P4
P40
1

P114
P113

P4

P112

P4

P111

P109
P108

Existing
basement
extent

P4

P4

47

0m
120

P107

P4
800 mm panels

P110

Existing caissons

P105

ls

SMA tunnel
protection zone

P104

P4

ane
mp

P106

P4

Outrigger
and vertical
anchors section

P5

1
P5
2

P5

P103

P5

P102

00
5

P6

P6

P6

P6

P6

P6

P7

P7

P7

P7

4
P7

P7

1200 mm panels

P7

P7

P7

P7

P8

P8

P8

P8

P8

P8

P8

P
P9 96
P9 5
4
P9
3
P9
2
P9
1
P9
0
P8
9
P8
8
P8
7

ls

P9

ne

pa

mm
800 mm panels
Bored panels

4
P6
3
800 mm
P6
2
P6 P6 panels
P P P P
1 0 59 58 57 56

P5

12

P
P 10
P 10 1
P9 99 0
8

P5

Geotechnical
section
(refer Figure 3)
N
0

10 20 30 40 50 60 70 80 90 100 m

Anchors

Figure 4. Diaphragm wall panel layout

(a) limiting lateral wall deflections to less than H/200, where H

is equal to the excavation height
(b) limiting settlement at the diaphragm wall toe to less than
0.3% of the panel thickness
(c) design of permanent support elements for a 100-year design life
(d ) design of the diaphragm wall that forms the basement and to
2 m below the basement floor to the requirements of Section
3.0 of AS 3735-2001 (Standards Australia, 2001): Concrete
structures for retaining liquids
(e) diaphragm wall to be substantially watertight with water
indications limited to minor damp patches on the wall surface
with no visible flow of water; groundwater ingress through
the diaphragm wall not to exceed 0.75 litres/min from the
entire constructed wall area enclosing the basement
( f ) building code of Australia importance level 4: Building or
structure that is essential to post-disaster recovery or
associated with hazardous facilities with an earthquake
annual probability of exceedance of 1:1500
(g) consideration of sea level rise due to climate change; this was
assessed to result in a groundwater level increase of 1 m over
the 100-year design life
(h) consideration of the effect of base and top slab shrinkage on
the diaphragm wall.
In order to satisfy these criteria, close collaboration between the

geotechnical and structural designers was essential to drive an

efficient reinforcement design that met the design expectations.
4.2 Key design parameters
Some of the key geotechnical parameters adopted for design are
summarised in Table 1. The soil and rock were modelled as
elastic-plastic material with a MohrCoulomb failure criterion in
the Plaxis analysis. The fill material was modelled using a
hardening-soil model, where the elastic unload/reload modulus
was three times stiffer than the primary compression modulus.
In the assessment of wallsoil interaction, both short-term and
long-term concrete Youngs modulus values were adopted in the
numerical analysis, with values of 20 000 MPa and 10 000 MPa
ascribed, respectively. For reinforcement design, a concrete
Youngs modulus of 34 800 MPa was used.
4.3 Watertightness
As the basement excavation will be below sea level, it is
essential that the diaphragm wall also acts as a temporary cutoff against groundwater ingress during construction. Once
constructed, the basement will be a fully tanked structure and
designed to accommodate full hydrostatic uplift pressures. A
minimum socket of 0.3 m into medium-strength or better rock
was adopted to control groundwater ingress from the under197

Geotechnical Engineering
Volume 167 Issue GE2

Barangaroo South harbourside basement,

Australia, challenges and solutions
Badelow, Wakefield, Chandrasegaran and
Iseppi

Surcharge 25 kPa

800th
panel
2

Anchor RL 120

Water level

Soil face

2
Fill
4

m RL 8
10

Basement RL 490
Excavation RL 585

gr Te
ou n s
nd tr
70 an and
s
0
c
kN ho tem
/m rs 1 p
pr 56 ora
r
es
tre m y
ss C/C

Excavation face

45

12
14
Toe RL 1620

16
CLASS III SST
18
20

CLASS II SST

Figure 5. Typical diaphragm wall section

Material

Fill
Estuarine sediments
Alluvial sediments
Residual soil
Weathered sandstone
Sandstone
a
b

Unit weight,
: kN/m3

Drained cohesion,
c9: kPa

Drained friction
angle, 9: deg

Elastic modulus,
E a : MPa

At rest earth
pressure
coefficient, K0

19
18
20
20
23
23

2
0
0
5
30
200300

30
25
31
25
35
35

1025
1025
2030
60
80
5001000

0.5
0.5
0.5
0.5
Variesb
Variesb

Elastic modulus values were selected considering the strain levels induced in the ground.
Estimated based on in situ horizontal stresses, which can be high in Sydney sandstone, with the major stress field in the NS direction.

Table 1. Summary of geotechnical design parameters

lying rock during construction. The embedment specification

was provided by the client and was based on careful review
of the geology, testing of the rock (including permeability
testing) and analytical evaluation of various depths of embedment.
198

4.4 Seismic design

Seismic loads acting on the diaphragm wall were assessed in
accordance with AS 4678-2002 (Standards Australia, 2002):
Earth-retaining structures. A pseudo-static analysis approach was
adopted using the MononobeOkabe method (Mononobe, 1924)

Geotechnical Engineering
Volume 167 Issue GE2

Barangaroo South harbourside basement,

Australia, challenges and solutions
Badelow, Wakefield, Chandrasegaran and
Iseppi

with the calculated pseudo-static loads applied to the diaphragm

wall in the finite-element analysis.

the wall axis and connected to the wall by a cast-in-situ concrete

outrigger beam (Figure 6). The vertical lock-off anchor force
generates a couple at the top of the wall, which in turn reduces
wall deflections. The anchor lock-off load was adjusted to
maintain a lateral deflection below H/200. The design was carried
out using subgrade reaction and finite-element analyses, both of
which indicated that anchors with a working load of 1400 kN at
1.6 m spacing would be required.

4.5 Space constraints

Along the southern site boundary, an existing basement located
just 9 m from the diaphragm wall precluded the use of inclined
ground anchors. To overcome this space constraint, the following
options were explored
(a) top-down construction, with installation of the ground-floor
slab prior to excavation
(b) raked-in struts supported on barrette portal frames installed
on the excavation side
(c) cantilever wall with vertical anchors.
Option (c) was preferred by the client as it did not impinge on
excavation sequencing. This solution comprised 1.2 m thick
diaphragm wall panels, embedded at least 3 m into mediumstrength or better sandstone. To control wall displacements,
vertical prestressed anchors were installed on the soil side behind
3000 m AHD

4.6 Future metro tunnels

The diaphragm wall had to include provisions for the construction
of the proposed Sydney metro tunnels, which cross the wall
alignment in the south-east corner and the western boundary.
In the south-east corner, the diaphragm wall was constructed
adjacent to and across the future tunnel alignment. The wall
panels were required to carry relatively large building loads as
the tunnel geometry restricted the use of piled foundations. The
design was further complicated by the presence of limited rock
cover above the tunnel crown. To transfer the building loads away
100 120
Cantilever capping beam

Temporary vertical anchors 10T152

at 16 m spacing
Lock-off load 1400 kN

Macquarie Bank
perimeter wall

5850 m AHD

13700 m AHD

Sandstone

Fill

300

Fill

Sandstone

Figure 6. Outrigger and vertical anchor arrangement

199

Geotechnical Engineering
Volume 167 Issue GE2

Barangaroo South harbourside basement,

Australia, challenges and solutions
Badelow, Wakefield, Chandrasegaran and
Iseppi

from the section of diaphragm wall which crossed above the

future tunnels, a number of the adjacent diaphragm wall panels
were extended to a greater depth to form a trouser-leg arrangement (Figure 7). These trouser legs were successfully constructed using a Hydrofraise to cut the high-strength sandstone.

(1998) using the volume loss concept. Using a weighted average

volume loss for the tunnels of 0.9%, as the tunnels will be partly
in alluvium and partly in rock, the resulting tunnel crown
settlement and lateral movement at the tunnel centreline were
estimated. Numerical analysis was completed to simulate the
tunnel excavation process and to obtain magnitudes of soil
movements that were similar to the values given by the analytical
methods. Greater emphasis was placed in matching the lateral
movement, as this has a greater influence on the wall panel
socketed into rock adjacent to the tunnel. Deflections and stresses
developed in the diaphragm wall due to tunnel excavation were
assessed from the numerical analysis and the reinforcement
design modified to accommodate the increased structural actions.

P39

P37

P38

The future tunnels also pass within close proximity of a number

of panels along the southern boundary, therefore the impact of
tunnel excavation-induced ground movements on these panels
had to be considered. The tunnelling-induced vertical and lateral
ground movements were first assessed using analytical methods
described by Mair et al. (1993) and Loganathan and Poulos

5.

Construction challenges and solutions

5.1 Varying rock levels and buried cliff lines

Where the rock was shallow (mainly along the eastern site
boundary), the diaphragm wall panels were terminated in rock
0.5 m below the excavation level and shear pins installed at their
base to provide adequate toe stability (Figure 8). The shear pins
consist of 127.5 mm diameter high-grade steel tubes embedded
2 m into the rock. Ducts, 200 mm in diameter, were fixed inside
the reinforcement cages to allow the drilling of the shear pins
through the completed panel. As these panels required considerable excavation into rock, a Hydrofraise Evolution 3 was used to
increase production rates and reduce vibrations that would have
resulted from conventional chiselling.
In areas where bedrock levels were deep (e.g. along the western

P41

At the tunnel crossing on the western boundary, the tunnel

crown and the toe of the diaphragm wall are separated by a
minimum of 4 m of sandstone. Finite-element analysis showed
that the vertical loads applied on the diaphragm wall in this area
will have negligible impact on tunnel excavation and lining
design. Therefore, no specific measures apart from the installation of removable anchors had to be implemented for this
section of the diaphragm wall.

P40

The future tunnels will be constructed through several diaphragm

wall panels, therefore soft eyes using fibre-glass reinforcement
were incorporated in a number of reinforcement cages to
facilitate tunnel excavation. Also, as most of the temporary
ground anchors clash with the tunnel alignment, removable
ground anchors were installed. After the base and ground-floor
slabs are cast, these anchors will be de-stressed and the strands
removed from the ground.

Basement
transfer
slab

Rock level
Soft eye zone

Future
tunnel
0 1 2 3 4 5 6 7 8 9 10 m

Figure 7. Diaphragm wall trouser-leg arrangement

200

Tunnel
reserve

Future
tunnel
Trouserleg
panels

Geotechnical Engineering
Volume 167 Issue GE2

Barangaroo South harbourside basement,

Australia, challenges and solutions
Badelow, Wakefield, Chandrasegaran and
Iseppi

Surcharge 25 kPa

800th
panel

Toe RL 250

Class III SST

m RL 4

Rock bolt RL 150

10

32 rock bolts
at 24 m centres

re
e

Fill

Th

Anchor RL 120

Excavation face

Water level

st
22 gro ran
5 un ds
kN d
t
/m anc emp
h
o
pr
es ors rary
tre
ss
Soil face

Shear pins 127 125 mm

at 16 m centres 20 m
into Class III SST or better

Class II SST

10

RL 1145
12

Figure 8. Diaphragm wall in areas of shallow rock

and northern boundaries of the site), the diaphragm wall panels

were constructed with a minimum socket of 0.3 m into mediumstrength or better rock. The rock socket provided both groundwater cut-off and bearing capacity. In a number of locations, the
stepped nature of the rock surface resulted in rock level changes
across the length of an individual panel. Where this occurred,
steps were introduced into the panel such that each bite achieved
the minimum embedment depth. As each panel was excavated in
three bites, a maximum of two steps were introduced along the
length of any one panel. At the peak of productivity, four
mechanical grabs were simultaneously operating on site and the
rock sockets were formed using grab and chiselling techniques.

5.2 Crossing of the existing caisson harbour wall

In the north-western corner of the site, the diaphragm wall
crosses the existing caisson harbour wall. These caissons are
reinforced concrete elements, each some 16 3 9 m in size and
15 m high, and composed of six cells infilled with sand and
separated by 0.20.3 m thick walls, with a 0.8 m thick raft
foundation founded on a gravel mattress.
The caissons had to be removed locally using rotary coring
techniques to allow for the 1.2 m thick diaphragm wall to be
constructed. Ten secant cores (1.5 m diameter at 0.9 m spacing)
were required to form a slot greater than the diaphragm wall
201

Geotechnical Engineering
Volume 167 Issue GE2

Barangaroo South harbourside basement,

Australia, challenges and solutions
Badelow, Wakefield, Chandrasegaran and
Iseppi

panel thickness, taking into account positional and verticality

tolerances. Once a core was complete, the casing was gradually
extracted and the hole backfilled with a mixture of clean sand
and cement. During these operations, caisson movements were
monitored on both sides of the trench using survey targets, and
monitoring records showed that the displacements remained
below threshold values.

In the north-western corner of the site, installation of ground

anchors was greatly complicated by the presence of the existing
caissons behind the diaphragm wall. Installation of the anchors
required drilling through several thick, reinforced concrete caisson walls. Buried steel beams and sheet pile walls were among
other subsurface obstructions encountered and sonic drilling
techniques were successfully used in these areas.
The depth of the bedrock in this area (over 30 m below ground
level) resulted in very long anchor free lengths (up to 50 m). As
the hydraulic jack stroke was limited and not sufficient to absorb
all the free length elasticity, a two-stage lock-off procedure was
required to perform the acceptance tests. Wedges were locked at
an intermediate load step, so enabling the ram to be pulled back
to its original position.

5.3 Anchor installation and stressing

The diaphragm wall panels are anchored by a single row of
temporary multi-strand ground anchors with a service life of
2 years. A total of 454 anchors were installed at an angle of 458
and bonded into 150 mm dia. boreholes drilled into mediumstrength or better sandstone. The typical working capacity for
these anchors ranged from 360 kN to 1700 kN, necessitating the
use of as many as 12 strands, each strand being 15.2 mm dia.,
plastic sheathed and greased along their free length. Anchor
reservations were built in the panel reinforcement cage, with
appropriate horizontal reinforcement provided such that an
external waler was not required.

5.4 Removable anchors

As excavation of the future metro tunnels would be hindered by
the presence of anchor strands, all temporary anchors had to be
removed in the tunnel protection zones. Fully removable anchors
were installed using methods based on single-bore, multipleanchor (SBMA) technology. SBMA consists of a number of short
unit anchors bonded to the rock at staggered depths in a single
borehole (Figure 9). Each anchor unit is equipped with a 1808
looped strand, greased and sheathed over its entire length, and
attached to a steel saddle installed in the borehole. The anchor is
prestressed by pulling on the strands at both ends and load is

Depending on the design cross-section, the temporary anchors

were locked off at a prestress load of between 60% and 90% of
the design working load in order to control wall lateral displacement. The lock-off procedure was in accordance with British
Standard BS 8081 (BSI, 1989).
Bond length of
single-bore multiple anchor

Sheathed greased
strands
Section AA

Grout

One or two looped strands

per unit anchor
Strands to other
unit anchors
Plastic sheaths or similar
around individual strands
Debonded strands to lower
unit anchors
(a)

Temporary removable single-bore multiple-anchor system

Grout

Unit anchor

(b)

Figure 9. Single-bore multiple anchor: (a) detail of single-bore

anchor; (b) section through single-bore multiple anchor

202

Unit anchors stressed simultaneously only by use of

hydraulically synchronised multiple jack system which
accommodates different unit anchor extensions

Geotechnical Engineering
Volume 167 Issue GE2

Barangaroo South harbourside basement,

Australia, challenges and solutions
Badelow, Wakefield, Chandrasegaran and
Iseppi

transferred from the unbonded strand to the saddle. A steel thread

bar fixed to the saddle transfers the load to the grout and thus to
the surrounding rock. When required, the greased, unbonded
strand can be removed by pulling on one end using either a winch
or a jack.

In the south-east corner, the diaphragm wall forms a 908 angle,

orientated towards the excavation, therefore the risk of cutting
an installed anchor during the drilling for other anchors was a
major concern. To mitigate the risk, the anchor spacing was
increased and anchors were installed at different inclinations
and skew angles to ensure that the distance between anchor
holes was always safe. In addition, sections of the diaphragm
wall had to be left anchor free to allow for the installation of
future piles behind the wall prior to the removal of the
temporary anchors.
The drilling of some anchors was hindered by the presence of old
timber piles that once supported wharf structures. Where such
obstructions were encountered, sonic drilling techniques proved
successful.

6.

Monitoring of diaphragm wall

performance

The performance of the diaphragm wall is being monitored by

inclinometers installed within the reinforcement cages and extending some 4 m below the panel toe and survey targets located
on the diaphragm wall (top, mid-height and base of excavation).

30

Lateral wall deflection: mm

20 10
0
10
20

30

40

2
4

Diaphragm wall depth: m

Altogether, 35 removable anchors were successfully installed in

the south-east corner and on the western boundary. Depending on
the capacity required, between four and 11 units were installed in
a 180 mm dia. borehole and each anchor was proof tested to 25%
above the working load. On completion of the load test and
acceptance of the anchor, each unit was individually reloaded and
locked off to 10% above the prestress load to account for friction
losses and wedge draw-in.

40
0

6
Excavation level
8
10
12
Measured
14

Predicted

16
18
20

Figure 10. Measured compared with predicted diaphragm wall

deflections

challenges. Some of the challenges faced required innovative

design solutions, including vertical anchors, removable anchors
and trouser-leg panels, as well as detailed numerical analysis to
carefully assess impacts of the diaphragm wall on future metro
tunnels, and vice versa.
Some of the project achievements described in the paper are
summarised below.

At the time of writing, the formation level has been reached and
lateral deflections remain within acceptable limits as defined by
the client. Generally the recorded lateral deflections into the
excavation are below the predicted movements inferred from the
design calculations and provide a level of confidence that the
diaphragm wall is performing as per the design intent. An
example plot for an inclinometer installed in the western
(harbourside) wall, where the formation level has been reached
but permanent support is yet to be installed, is given in Figure 10,
together with the deflection profile predicted from the Plaxis
analysis for the diaphragm wall at this construction stage for
comparison.

Watertightness was achieved by limiting the number of

anchors used below groundwater level and reinforcement
design to AS 3735-2001 (Standards Australia, 2001).
j Groundwater cut-off was achieved economically with a
minimum wall penetration of 0.3 m into medium-strength or
better sandstone.
j Detailed assessment was undertaken of the diaphragm wall
impacts on the future metro tunnels, and conversely the
impacts of tunnelling-induced ground movements on the
diaphragm wall performance.
j Challenging conditions for construction, including cliff
ledges, building over future metro tunnels and space
constraints due to adjacent basements, were dealt with by
detailed geotechnical testing/monitoring and smart design.

7.

REFERENCES

Conclusion

The collaborative approach adopted by the design and construction teams resulted in meeting a very tight delivery programme
for the diaphragm wall while overcoming significant in-ground

BSI (1989) BS 8081: British standard code of practice for ground

anchorages. BSI, London, UK.

Loganathan N and Poulos HG (1998) Analytical solutions to

203

Geotechnical Engineering
Volume 167 Issue GE2

Barangaroo South harbourside basement,

Australia, challenges and solutions
Badelow, Wakefield, Chandrasegaran and
Iseppi

predict tunnelling induced ground movements. Journal of

Geotechnical and Geoenvironmental Engineering, ASCE
124(9): 846856.
Mair RJ, Taylor RN and Bracegirdle A (1993) Subsurface
settlement profiles above tunnels in clays. Geotechnique
43(2): 315320.
Mononobe N (1924) Consideration into earthquake vibrations
and vibration theories. Journal of Japan Society of Civil
Engineering 10(5): 10631094.

Standards Australia (2001) AS 3735-2001: Concrete structures

for retaining liquid. Standards Australia, Sydney, NSW,

Australia.
Standards Australia (2002) AS 4678-2002: Earth-retaining
structures. Standards Australia, Sydney, NSW, Australia.
Wong P, Bressi R, Jerogin P and Kardash V (2013) From
wasteland to world-class centre: Barangaroo South, Sydney.
Proceedings of the Institution of Civil Engineers Civil
Engineering 166(5): 39.

WHAT DO YOU THINK?

To discuss this paper, please email up to 500 words to the

editor at journals@ice.org.uk. Your contribution will be
forwarded to the author(s) for a reply and, if considered
appropriate by the editorial panel, will be published as a
discussion in a future issue of the journal.
Proceedings journals rely entirely on contributions sent in
by civil engineering professionals, academics and students.
Papers should be 20005000 words long (briefing papers
should be 10002000 words long), with adequate illustrations and references. You can submit your paper online via
www.icevirtuallibrary.com/content/journals, where you
will also find detailed author guidelines.
204