Вы находитесь на странице: 1из 10

CENTRAL SAINT GILES BASEMENT AND FOUNDATIONS DESIGNED FOR

FUTURE CROSSRAIL TUNNELLING


Chris Barker, Marek Niewiarowski and Dinesh Patel, Arup, London, United Kingdom
In London, it is more common to design the foundations and substructure of new
developments adjacent and over existing tunnels. However, designing economic
substructures for future tunnelling presents a set of unique design challenges
resulting in bespoke solutions tailored to the structure, ground and severity of the
imposed tunnelling constraints.
This paper presents several geotechnical
challenges in the design of a basement structure that cantilevers out over the
safeguarded Crossrail tunnel corridor and the foundations which need to
accommodate ground movements induced by this future tunnelling. The design and
construction outcomes for the basement foundation scheme are discussed and
comments are made on construction observations, pile testing and monitoring.

INTRODUCTION
The Central Saint Giles development is located in
the West End of London where future Crossrail
tunnels will be constructed. These tunnels will be
sufficiently close as to impact on the basement and
foundations for this new development.
Crossrail is a major proposed east-west railway
across London. Since 1990, the route alignment
has been safeguarded through the local authority
planning application process to ensure future
construction of Crossrail is not prejudiced by any
other development.
Crossrail construction is
planned to commence in 2010 with services
starting in 2017.
Through central London, fifteen kilometres of
seven metre diameter twin bored tunnel will weave
its way between existing railway and other tunnels,
sewers and building foundations. At Central Saint
Giles, the eastbound tunnel alignment passes
beneath the southern boundary of the site. The
tunnel crown is about 16.5m below ground level
with a 4m exclusion zone above and 3.5m either
side of the tunnel to allow Crossrail flexibility to
move the proposed tunnel during design.
It is mandatory that all new projects within the safe
guarded zone must consult Crossrail. At Central
Saint Giles consultation with Crossrail commenced
during concept design stage where tunnel location,
geometry, loading and settlement constraints were
advised by Crossrail for which the design of the
Central Saint Giles development had to address in
design. These constraints had a defining influence
over the foundation and substructure solution as
discussed below.
CENTRAL SAINT GILES DEVELOPMENT
Central Saint Giles is a striking new landmark in
Londons West End just east of Centre Point Tower

by architect Renzo Piano.


The mixed use
development comprises a fifteen storey
reinforced concrete frame residential building and
an eleven storey steel frame office building set
around a central courtyard. The superstructure is
clad by dramatic facades of vivid colours red,
orange, green and yellow glazed ceramic which
sit on the perimeter basement wall, Figure 1.

Figure 1: Impression of development (Legal &


General)
This eye-catching development is supported on a
site wide seven metre deep single level
basement substructure which, with the use of
900mm thick full basement height transfer beams
along the southern frontage, overhangs the
proposed eastbound Crossrail tunnel exclusion
zone by up to 8.5m, see Figure 2. A local three
metre deep second level basement is located in
the east of the site.
Demolition of the pre-existing building and
backfilling the site was completed by Keltbray Ltd
in late 2007 with Stent Foundations Limited
following on immediately with preliminary pile

Figure 2: Foundation layout with 14m wide


Crossrail tunnel exclusion zone
testing and piling which was completed in April
2008. Building frames were completed in July
2009 and the Stanhope Plc development is due for
final completion in March 2010.
THE SITE AND GROUND CONDITIONS
The large city site is irregular in shape with
maximum dimensions 280m east-west and 190m
north-south. Surrounding street levels are
+25mOD in the north falling to +23mOD in the
south east. The previous building on site was a
1950s seven to nine storey S shaped masonry
Ministry of Defence (MoD) building with a single
level basement founded on shallow footings.
A ground investigation of twenty-one geotechnical
and archaeological trials pits, three cable
percussive with follow on rotary boreholes and
twelve concrete cores at the site encountered
variable thickness of Made Ground up to 4.5m
deep which included numerous truncated and
backfilled historic basements and foundations predating the MoD building. Underlying this Made
Ground was about 2.5m of Terrace Gravels then
only 21m of London Clay. Cable percussive
boreholes where then continued by rotary coring to
prove 17.5m of Lambeth Group clays and a thin

6m horizon of Thanet Sands overlying the


bedrock Chalk proven to a couple of metres
depth.
Ground water comprises a shallow aquifer some
4m below ground level (up to 1.8m above the
surface of London Clay) and a depressed deep
aquifer 12m below the top of the chalk with the
overlying London Clay, Lambeth Group
(comprising Upper Mottled Clay, Laminated
Beds, Shelly Clay, Lower Mottled Beds and
Upnor Formation) and Thanet Sand strata being
underdrained.
SUBSTRUCTURE DESIGN
The proposed eastbound Crossrail tunnel
beneath the southern boundary had a defining
influence over the foundation and substructure
solution. Crossrail imposed the following design
constraints on the engineers designing the new
building:
outside tunnel diameter of 7m;
allowance for the advised tunnel alignment to
move up to 3.5m horizontally and 4m
vertically;

loading imposed on the tunnels does not


exceed existing overburden plus loading from
existing development (or plus 50kPa at ground
surface);
design of foundations for the development to
allow for 1.7% face loss from tunnelling; and
operational noise and vibration values were
also provided.
While a few substructure possibilities were
considered at concept design stage, the chosen
solution was for a bottom up basement supported
on piles and cantilevering out over the tunnel. The
piles being designed for future tunnelling induced
ground movements and the excavation formed with
a load bearing perimeter embedded retaining wall
with temporary propping.
To carry faade column loads up to 5.5MN, the
load bearing perimeter embedded wall was
1180mm diameter (reducing to 1050mm diameter
below casing at +15.5mOD) hard-firm rotary bored
secant wall at 980mm centres with between 1.7%
and 2.6% steel. At about 10m from the tunnel
exclusion zone, it was continued as a propped
temporary 600mm diameter CFA hard-firm secant
wall to allow the cantilever sub-structure to be built.

The permanent southern boundary basement


wall was a 900mm thick cast in situ reinforced
concrete beam and was cantilevered with eight
900mm thick full basement height transfer beams
tied back into pile caps, basement and ground
floor slabs as shown in Figure 3. The basement
slab through this zone including pile caps was
underlain with compressible void former to
maximise the build out of basement settlement
during construction. Structural loads (excluding
downdrag allowance) on southern pile caps
typically ranged between 20MN to 35MN with the
western concrete framed residential building pile
cap carrying 110MN.
A prediction of ground movements resulting from
future Crossrail tunnelling was made using the
Oasys computer program, TUNSET. Based on
Crossrails anticipated face loss of 1.7% from
construction of running tunnels, ground
movements above the tunnel in front of the
basement were predicted to be in the order of
30mm to 50mm between ground level and tunnel
crown. Also a 45 degree wedge extending up
from the tunnel to basement formation level
defined a zone of predicted tunnelling induced
ground settlements greater than 5mm. This
demarcated a 6m wide zone from the edge of the
tunnel exclusion zone for piles to be designed for
downdrag.
The substructure was founded on 900mm and
1500mm diameter rotary bored bearing piles
founding in the London Clay and Lambeth Group
clays carrying loads between 1.6MN to 10.1MN.
However, to reduce the additional effect of
downdrag within the 6m wide downdrag zone,
piles were sleeved with a 12mm thick 1560mm
OD permanent steel casing coated with 6mm of
Bitumen Compound SL and installed and grouted
in an over size bore to one metre above tunnel
invert level. Piles within this zone were founded
about 47m below ground level, a metre into the
Thanet Sand with the twenty piles under the
residential pile cap also base grouted.
TRIAL BORES

Figure 3: Cross section through cantilever


basement

Of the 59 No. 1500mm diameter rotary bored


piles 35 were founded into the Thanet Sand and
16 others at various levels within the Lambeth
Group. It was anticipated that due to the basal
London Clay and Lambeth Group being
historically underdrained for a considerable
period of time seepages and instability would be
minimal. However, it was difficult to assess shaft
degradation in the relative small diameter
boreholes during the ground investigation.
Consequently, bentonite support fluid had been
allowed for at tender due to the known instability

in the Thanet Sand and sandy horizons of the


Lambeth Group and to a lesser extent the
remainder of the Lambeth Group and basal sandy
London Clay horizon.

Also significant spalling of the lower Lambeth


Group occurred over a basal 1.5m height. More
seepages were evident in the London Clay than
Lambeth Group.

Also within the tender scope, two trial bores with


CCTV recording were also specified as part of the
preliminary pile testing programme to identify the
specific formations within the bores that exhibited
degradation during a piling period of up to several
hours. This in turn allowed an assessment to be
made as to what stage in the boring process the
support fluid would need to be introduced, and to
what level the fluid would need to be retained
within the bore.

Auger boring of the dry Thanet Sand stratum was


then attempted and it was found that collapse
was occurring immediately upon withdrawal of
the auger with the result that after four auger
loads the bore still dipped at the top of the
Thanet
Sand.
CCTV
footage
showed
undercutting and slumping beneath the Upnor
Formation. Re-drilling resulted in the same effect
and it was considered that due to the dry
moisture conditions, the low suction in the Thanet
Sand was overcome by a piston effect of a fully
loaded auger being withdrawn which resulted in
bore collapse. While this could have been
overcome by using a vented digging bucket or
double flight auger, the significant spalling of the
lower Lambeth Group was already a deciding
factor. This has also been observed by the
author more recently on the Canary Wharf
Crossrail Station project, again during trial boring
in the Thanet Sand stratum.

The trial bores were constructed using a Soilmec


R-625 rotary piling rig at 900mm and 1050mm
diameters, which allowed free movement and
rotation of the CCTV camera during subsequent
monitoring and was sufficiently large to observe
bore degradation in the large diameter piles being
adopted for the site. After installation of temporary
casing into the London Clay, boring without
support fluid was carried out uninterrupted through
the London Clay and Lambeth Group to the top of
the Thanet Sand. A record of the material
encountered during excavation was made by the
Resident Engineer and CCTV monitoring
commenced immediately upon the completion of
boring.
CCTV monitoring of the type more commonly used
in drainage surveys, was carried out over a
continuous period of about 26 hours using a colour
survey camera lowered into the bore. The camera
was mounted with strong lighting and had 360
degree turn and tilt functions which were remotely
controlled from a monitoring van at the surface.
Images were continuously recorded onto DVD.
After completion of boring, the camera was
lowered to the base of the excavation, recording
the location and extent of any seepages or
degradation of the bore surface along the way. A
slow process of travelling up and down the bore
continued until the first major degradation was
observed in sandy beds near the base of the
Lower Mottled Beds four hours into monitoring
after which the camera was then left to record the
development of the major bore degradations or
breakouts with up and down surveys to observe
the development of new breakouts undertaken
approximately every 30 minutes.
Figure 4 summarises the location and extent of
seepages and degradation recorded over the
monitoring period for Trial Bore 1. Degradation of
the pile shaft continued steadily with small
fragments of soil raining to the base of the pile.

Figure 4: Degradation & seepage in trial bore

From the trial bore observations it was concluded


that bentonite was not required to be introduced
during boring through the basal London Clay or
Lambeth Group strata, but would be introduced
into the pile bore upon reaching the readily
identifiable flint marker beds in the Upnor
Formation immediately above the Thanet Sand.
This enabled considerable programme benefits to
be realised with the number of scheduled bentonite
bored piles being reduced from 51 to 35. This
reduction was achieved by being able to recalculate a higher pile toe level, above the
observed trial bore pile instabilities, using a higher
non-bentonite shaft adhesion alpha value.
TEST PILES
Two preliminary pile tests were carried out in
August 2007, approximately one month in advance
of the main contract pile works. Preliminary test
pile PTP1 was specified to confirm assumed
design basis for piles founded entirely in London
Clay. Preliminary test pile PTP2 was specified to
demonstrate settlement performance of pile end
bearing in the Thanet Sand and to measure
bitumen coated shaft adhesion to minimise
calculated downdrag loads. Two working test piles
were carried out in April 2008 during the main
piling works, as required by the London District
Surveyors Association guidance, LDSA (2000) for
the factor of safety adopted.
PTP1 was a 900mm diameter straight shafted
rotary bored pile founded entirely in London Clay.
The design philosophy followed the LDSA
guidance notes for the design of straight shafted
piles in London Clay. Table 1 of the LDSA
guidance note suggests a FoS of 2.0 and shaft
adhesion factor, = 0.5, for a Maintained Load
(ML) test. The pile test was chosen to be a ML test
since these better reflect pile-loading conditions
and are less susceptible to rate effects which in
stiff clay can lead to high excess pore pressure
effects. The location near to BH3 (see Figure 2)
was selected taking into account the proposed
permanent works pile locations, location of ground
investigation boreholes, basement demolition
works and archaeological investigations being
undertaken on site. Double sleeved casing was
installed to minimise the maximum test load by
isolating the pile shaft through the Made Ground
and River Terrace Deposits and also to negate the
risk of structural failure of the pile head under load.
Pile construction was independently witnessed by
the Resident Engineer. No notable difficulties were
encountered during construction. The pile bore
and base were clean and dry, and free of
seepages. Boring and base cleaning of the pile
was uninterrupted and took approximately 1.5

hours. The duration from base cleaning to the


start of pile concreting was approximately 2.5
hours. Concreting of the pile was uninterrupted
and took approximately 1.25 hours. Concrete
cube test and sonic logging results were
satisfactory.
PTP1 was instrumented with 4 levels of strain
gauges, with extensometers fitted near to the pile
base and 0.5 m below the bottom of the double
sleeving. Electronic displacement transducers
measured movement at the head of the pile.
Precise levelling of the top of the pile head was
also carried out as a check on pile head
movements. The sister bar type strain gauges
were used to analyse the load shed down the
length of the pile, from which the design
parameters could be back-calculated. One
Geokon retrievable multi-point extensometer was
installed in the pile for the duration of load
testing. The extensometer was divided into two
sections through the use of three pneumatic
anchor points linked by displacement transducers
and variable lengths of steel rod. All of the strain
gauges appeared to be recording throughout the
test.
PTP1 was tested by a maintained load test in
25% increments to 100% DVL, followed by an
extended proof load test to DVL + 50% SWL. It
was then intended to follow this with a
maintained load test to 200% DVL. However, the
maximum load achieved in the maintained load
test was 168% DVL. In order to calculate the load
transfer occurring along the length of the pile, a
relationship between the pile secant modulus and
measured strain was calculated using the
tangent stiffness method described by Fellenius
(2001).
The pile behaved in an elastic manner during the
first loading up to 100% DVL (3100 kN).
Settlement at this stage was 3.3mm which is
comfortably within the 4 mm maximum set in the
specification after Patel (1992). On loading up to
125% of DVL the pile continued to behave
elastically, the majority of the load being carried
by shaft friction with little transferred to the base.
When 150% DVL was applied the settlement was
0.5mm in the first hour. This load was maintained
for 36 hours during which the settlement declined
slowly but erratically. The settlement rate was
0.18mm per hour in the last half hour reaching a
maximum settlement of 19.2mm. The pile was
then unloaded and reloaded to 150% DVL. The
settlement rate was over 1mm/hr in the first hour.
The load was maintained for 14 hours during
which the rate diminished, again erratically, to
0.2mm/hr in the last 30 minutes. An attempt was
made to raise the load to 175% DVL (5425 kN),

but the pile settled 50mm in 20 minutes and the


load was not reached.

programming and to keep the maximum test load


within limits of readily available load frames.

Back analysis on the basis of the design clay


strength profile and an overall factor of safety FoS
= 2.0 (as per the LDSA approach for ML tests)
showed that the mobilised shaft adhesion factor
peaked at = 0.4 at 129% of DVL after this the
shaft friction dropped and the load was transferred
to the base. This alpha value is lower than the
alpha of 0.5 suggested in the LDSA guidance
notes (LDSA, 2000). Possible causes investigated
included an error in the test process (calibrations,
etc), problems during construction, and the clay
strength profile. However, no unusual difficulties or
delays were experienced during construction, and
the clay strength profile was reviewed in detail and
was considered appropriate for the site. A review
of previous London Clay pile tests showed that the
performance, measured by , was within the range
of previous London tests but below the 0.5 value
recommended by the LDSA. It is noted that the
LDSA guidance alpha values are not lower bound
values and very occasionally alpha values below
0.5 are back calculated from preliminary pile tests.
For the detailed design of piles founded entirely in
the London Clay it was preferred to use the LDSA
parameters for Maintained Load (ML) tests, that is
FoS = 2.0, but to reduce the value from 0.5 to 0.4
to account for the low mobilisation of shaft friction
in the PTP1 and the existing design mean clay
strength profile of cu = 90 + 8.0z kPa was retained.

PTP2 was instrumented in a similar fashion to


PTP1, in this case with 10 levels of strain gauges
and 3 levels of retrievable multi-point
extensometers, as shown in Figure 5.

Two 900mm diameter working test piles founded in


the London Clay were carried out to give
confidence in the performance of the pile
foundations and revised design basis.
Both
working test piles were successfully tested to the
specified loads (i.e. up to 100%DVL + 50%SWL)
with acceptable settlements of less than 4mm at
100%DVL and less than 9mm at 100%DVL +
50%SWL. It was concluded that the design method
adopted was sufficient to ensure the margin of
safety for the piles.
Preliminary Test Pile 2 (PTP2) was specified to
demonstrate the performance of a pile constructed
through the London Clay and Lambeth Group,
bearing in the Thanet Sand, constructed under
bentonite support fluid. More specifically, the test
was intended to confirm:
shaft adhesion through the Lambeth Group;
a minimum end bearing factor in the Thanet
Sand; and
back calculation of shaft friction for bitumen
coated sleeving.
PTP2 was not base-grouted. A 1200mm diameter
test pile was chosen to provide both a practical
construction trial of this unusual pile for

Figure 5: Diagrammatic
arrangement, PTP2
The PTP2 construction
summarised as follows:

cross

sequence

section
can

be

From ground level install 1860mm diameter


temporary casing into top of London Clay
Bore 1800mm diameter to the design level
for friction reduction bitumen coated sleeving
Install 1260mm diameter bitumen slip coated
permanent sleeve to base of open bore,
having checked pre-applied bitumen coating
for consistency of coverage
Check position, level and verticality of
sleeve, and grout annulus between open
bore and permanent sleeve
Bore 1200mm diameter to approx 1m above
top of Thanet Sand

Introduce bentonite and continue boring to


target toe depth using a bypass digging bucket

Settlement at Pile Head (mm)


0

Clean base of pile using bypass cleaning


bucket. Check hardness of base using
weighted tape
Exchange bentonite in the bore with fresh
bentonite (not undertaken in works piles)

PTP2 was tested by a maintained load test in 25%


increments to 100%DVL, followed by an extended
proof load test to DVL + 50% SWL, followed by a
maintained load test to 225% DVL. The test was
ceased at 23.7MN (225% of DVL) with a pile head
settlement of 79mm and a rate of settlement within
the specification limit of 0.24mm/hr. As with the
back analysis of PTP1, the tangent stiffness
method described by Fellenius (2001) was used to
calculate the load transfer occurring along the
length of the pile. The measured load vs pile head
settlement is shown in Figure 6 and summarised in
the table below, along with some other similar
deep bitumen sleeved preliminary test piles
undertaken in London.
Site
Central Saint
Giles
Moorhouse
Pinnacle

Pile
Diameter
(mm)

Pile
Length
(m)

Length
Sleeved *
(m)

Base
Grouted

1200

47

7 / 16

No

900

56

26 / 22

Yes

900

64

11 / 40

Yes

Kings Cross
1200
37
0 / 25
Yes
CTRL C105
* Pile length sleeved: air void (i.e. double sleeved) length /
bitumen coated length.

These central London preliminary test piles were


all founded into the Thanet Sand. While pile
diameter and amount of sleeving differs across the
sample, Figure 6 illustrates pile performances
achievable with deep large diameter sleeved piles
in London.

Load at Pile Head (MN)

Satisfactory base hardness checks demonstrated


the base to be Grade 2 Firm prior to concreting,
using the base hardness scale (of 1 to 5) for
Thanet Sand piles constructed under support fluid.
Checks were also made on the minimum 6mm
thickness of bitumen slip coating material (prior to
installation of the permanent sleeve); bentonite
support fluid was sampled and tested for density,
viscosity, sand content and pH; concrete cubes
were sampled and tested for strength; and sonic
logging was carried out to confirm the integrity of
the pile.

50

75

100

20

Test bentonite and base hardness and install


reinforcement cage
Check base hardness again using weighted
tape, and concrete pile by tremmie.

25

25

15

10
Central Saint Giles
Moorhouse

Pinnacle
King's Cross CTRL C105

Figure 6:
comparison

Test

pile

load

settlement

As noted by Whitworth et al. (1993), the use of


bitumen applied as a slip coating to driven piles
has been investigated. According to Laybond
(1989), load transmission through Bitumen
Compound SL caused by moving soil transmitted
through bitumen coating is a function of:
Temperature of coating
Rate of settlement or movement of soil
Thickness of coating
Flow characteristics of coating under load
However, published data from bored test pile
case histories of bitumen coated permanent steel
cased piles are much less common. Whitworth et
al. (1993) reported the results of comparison pile
testing on coated (6mm bitumen) and non-coated
12m deep 750mm diameter pile tests at Angel
Square, London. At maximum test load of 325kN
at 17mm head displacement (2% pile diameter) a
still increasing average bitumen fs value of 11kPa
can be determined.
From the Central Saint Giles back analysis of
three sets of strain gauges the instrumentation
installed for the bitumen coated sleeving, the
results shown in Figure 7 indicate a rapidly
increasing initial average sleeve friction which
perhaps peaked at 15kPa then reduced with
average shaft movement over the length of
sleeve.

125

Lambeth Group to the maximum verified value of


100kPa.

Displacement (mm)
0

10

20

30

40

50

60

Similarly, in the Thanet Sand the end bearing


factor value of Nq was shown to be continually
increasing with load up to the maximum pile test
load, and therefore a limiting value for Nq of 13
was considered appropriate for design purposes
(no base grouting).

20
Bitumen sleeve friction (kPa)

18
16
14
12
10
8
6
4
2

Central Saint Giles (average


pile shaft movement)

Individual 1500mm diameter Thanet Sand rotary


bored
working
piles
were
successfully
constructed over a period of four days. The
sequence of construction for a typical pile at
Central saint Giles is shown in the table below.

Angel Sq, Islington (pile head


displacement)

Figure 7: Bitumen coated sleeve adhesion (fs)


After consideration of the expected range of
tunnelling induced ground settlement for the
bitumen coating, a design skin friction, fs, of 20kPa
was adopted in calculations of downdrag acting on
bitumen coated piles.
During the programme of preliminary pile testing it
was observed that the black bitumen slip coating
compound was found to melt and run during the
warm summer weather (~30C). However, this
was not found to be a problem during the main
piling works, which were carried out over the winter
period. Methods of controlling bitumen run during
hot periods included shielding the bitumen coating
from direct sunlight with tarpaulins and sheets, and
regular rotation of the sleeves, which were being
stored horizontally on timbers.
With a large proportion of the London Clay being
potentially mobilised during tunnelling it was
important to maximum shaft capacity through the
Lambeth Group and confirm this by preliminary pile
testing. Case history reporting of shaft friction
values in the Lambeth Group suffer from significant
variation reflecting the variable nature of this
stratum. Table A3.4 in CIRIA C583 reports fs
values such as: 84kPa Canary Wharf; 66kPa
Euston; 75kPa-113kPa British Library; 185kPa
Blackwall Yard and at a recent project at Kings
Cross 146kPa was back calculated from test piling.
In the Lambeth Group at Central Saint Giles, a
shaft adhesion factor of = 0.32 was back
calculated and adopted for design. Although PTP2
was constructed under bentonite support fluid, a
shaft adhesion factor of = 0.4 was again back
calculated for the short section of shaft in the
London Clay. However, it was decided to limit the
design shaft adhesion in both the London Clay and

Day
1

Activity
Install 1860mm diameter temporary
casing into top of London Clay
Bore 1800mm diameter to the design
level for friction reduction bitumen
coated sleeving
Install 1560mm diameter bitumen slip
coated permanent sleeve to base of
open bore
Grout annulus between open bore and
permanent sleeve (stage 1)
Grout annulus between open bore and
permanent sleeve (stage 2)
Bore 1500mm diameter to bentonite
introduction level
Introduce bentonite
Bore under bentonite and clean pile
base
Exchange bentonite
Install reinforcement cage and concrete
tremmie
3
Concreting (approx. 78m )

Typical
duration
(hours):
0.75
0.75
2.00
0.50
0.50
1.75
1.00
0.75
Not
required
4.00
4.00
16

RETAINING WALLS
In recent times the demands on commercial and
residential basements at least in London, is for
increased space for mechanical and electrical
plant necessary to also meet sustainability and
planning requirements as well as commercially
optimising roof space and appearance, the
traditional location for some plant.
Hence
basement heights have been on the increase so
that the single level seven metre basement
height at Central Saint Giles to facilitate plant
items and their maintenance is now not
uncommon.
With the complexity and associated cost of the
cantilevered area of basement, the challenge
was to also maximise savings for the remaining
perimeter wall.
With large basements like
Central Saint Giles there is less pressure on
minimising wall size as there is on small

basements where basement space is a premium.


It was therefore more feasible to consider
increasing the secant wall pile diameter and
cantilevering the basement wall during construction
in order to benefit from unrestricted access during
basement and building core construction.
In the temporary condition, retained heights to B1
basement formation along the permanent secant
wall varied from 5.5m to 8.2m with some sections
of wall close to pile cap and B2 basement
excavations giving effective retained heights up to
8.8m. A review of central London case studies
showed limited published case histories of
cantilever basements above 6m.

and building core construction. The southern


boundary temporary CFA secant wall having an
embedment less than its retained height due to
the Crossrail exclusion zone required berms and
raking props with horizontal props set above
capping beam level at the corners until
construction of both basement and ground floor
slabs. The July 2008 progress photo in Figure 8
shows an easterly view the basement under
construction with temporary propping in place.

Nevertheless, CIRIA C580 Figure A2.1 indicates


cantilever wall deflections in excess of 0.6% of the
retained height.
A damage assessment was
undertaken to consider surrounding buildings,
buried utilities, footpaths and road pavements. The
nearest surrounding buildings were generally
about 10m from the perimeter wall and with one or
two levels of basement and therefore a
conservative assessment classified them within an
acceptable very slight to slight damage category
according to Boscardin and Cording (1989).
A review of surrounding utility services indicated
favourable conditions. Pressurised utilities which
were considered more susceptible to damage such
as gas were modern relatively flexible low pressure
2.5 and 5 polyethylene butt welded construction
and the existing mains water pipes were scheduled
to be replaced in 2007 with similar MDPE butt
welded pipes as part of the Thames Water mains
renewal programme prior to construction. Also, the
12 concrete and 48 oval masonry sewers were
set back beneath the middle of the surrounding
streets and relatively deep at between four to five
metres below street level.
The assessment indicated that generally for areas
of single level basement cantilever deflections
would be acceptable because the significant
settlements would be expected to occur close
behind the wall and as part of this development the
surrounding footpaths and Dyott Street where
being reinstated.
Therefore the temporary propping scheme for the
permanent secant wall was economised to a single
level of horizontal props at capping beam level in
the north east and north west corners of the site
where deeper excavations for the B2 basement
and residential pile cap approached the secant
wall and effective retained heights increased.
These props were removed after construction of
the basement slab (acting as base prop to
cantilever) to allow less restricted ground floor slab

Figure 8: Site progress photo July 2008 (Mike


O'Dwyer)
Instrumentation
and
monitoring
of
the
construction comprised reflective survey targets
on surrounding buildings, levelling studs on
footpaths, thirteen inclinometers in selected male
secant piles and reflective targets on the capping
beam. In addition, precise levelling points for
future monitoring were set in the basement
transfer beams at the request of Crossrail.
Monitoring of two cantilever inclinometers as
shown in Figure 9 gave maximum deflections
much less than case history data indicated.
Maximum total deflections of 16mm and 18mm,
being 0.25% and 0.23% of their retained heights,
of 6.5m and 7.8m respectively were measured.
This includes a creep rate of 1.5mm and 2.5mm
per month respectively. The duration from
reaching
maximum
formation
level
to
construction of basement slab was about six
months and ground floor slab followed about
three months later.

Preliminary instrumented pile testing proved a


London Clay alpha value of only 0.4 and
confirmed a maximum Lambeth Group adhesion
of 100kPa and bitumen friction of 15 kPa.
The perimeter basement wall was also
instrumented and monitored during construction
and confirmed maximum cantilever deflection of
0.25% H for retained heights up to 7.8m.
REFERENCES
BOSCARDIN, M.D., and CORDING, E.J., 1989.
Building Response to excavation induced
settlements, ASCE Journal of Geotechnical
Engineering, Vol 115, No.1, Jan 1989.
FELLENIUS, B.H., 2001. From Strain
Measurements to Load in an Instrumented Pile,
Geotechnical News Magazine, Vol. 19, No. 1,
pp.35-38.
GABA, A.R., SIMPSON, B., POWRIE, W.,
BEADMAN, D., 2003. Embedded retaining walls
guidance for economic design, CIRIA C580.

Figure 9: Cantilever wall deflection profiles


These cantilever deflections are plotted below with
CIRIA C580 and St John et al (1992) data in Figure
10 for comparison and to show that significant
cantilever heights are possible in urban
environments.

Maximum horizontal wall defection delta (mm)

70

LDSA, 2000. Guidance Notes for the Design of


Straight Shafted Piles in London Clay, London
District Surveyors association, October 2000.

0.6%H

Key
Central Saint Giles data
+ C580 & St John et al (1992) data

60

HIGHT, D.W., ELLISON, R.A., PAGE, D.P.,


2004. Engineering in the Lambeth Group. CIRIA
C583, pp 206-210.

50

LAYBOND PRODUCTS LTD, 1989 Bitumen


Compound SL, Slip layer for bearing piles and
buried structures, C1/SfB 18, June 1989.

0.4%H

40

ORIORDAN, N.J., 1982. The mobilisation of


shaft adhesion down a bored, cast-in-situ pile in
the Woolwich and Reading Beds, Ground
Engineering April 1982, pp 17-26.

30

0.2%H
20

Patel, D.C., 1992. Interpretation of results of pile


tests in London Clay, Piling Europe, Institution of
Civil Engineers, April 1992.

10
0

CONCLUSIONS

ST JOHN, H.D., POTTS, D.M., JARDINE, R.J.,


HIGGINS, K.G., 1992. Prediction and
performance of ground response due to
construction of a deep basement at 60 Victoria
Embankment, Proc Wroth Mem Symp, Predictive
Soil Mechanics, Oxford pp 581-608.

The Central Saint Giles basement has been


designed and constructed to cantilever out over
the proposed Crossrail alignment and to
accommodate future tunnelling induced ground
movements.

WHITWORTH, L.J., TURNER, A.J., LEE, R.G.,


1993. Bitumen slip coated trial piles and
prototype undeream trial pile at Angel Square,
Islington, Ground Engineering, January/February
1993, pp 28-33.

10

Maximum excavation depth below ground level H (m)

Figure 10: Embedded cantilever retaining wall


deflections after CIRIA C580 and St John (1992)

12

Вам также может понравиться