FOUNDATION DESIGN FOR THE BURJ DUBAI – THE WORLD’S TALLEST

BUILDING
Harry G. Poulos Grahame Bunce
Coffey Geotechnics Hyder Consulting (UK),
Sydney, Australia. Guildford, UK

ABSTRACT

This paper describes the foundation design process adopted for the Burj Dubai, the world’s tallest building. The foundation system is a
piled raft, founded on deep deposits of carbonate soils and rocks. The paper will outline the geotechnical investigations undertaken,
the field and laboratory testing programs, and the design process, and will discuss how various design issues, including cyclic
degradation of skin friction due to wind loading, were addressed. The numerical computer analysis that was adopted for the original
design together with the check/calibration analyses will be outlined, and then the alternative analysis employed for the peer review
process will be described. The paper sets out how the various design issues were addressed, including ultimate capacity, overall
stability under wind and seismic loadings, and the settlement and differential settlements.

The comprehensive program of pile load testing that was undertaken, which included grouted and non-grouted piles to a maximum
load of 64MN, will be presented and “Class A” predictions of the axial load-settlement behaviour will be compared with the measured
behavior. The settlements of the towers observed during construction will be compared with those predicted.

INTRODUCTION GEOLOGY

The Burj Dubai project in Dubai comprises the construction of The geology of the Arabian Gulf area has been substantially
an approximately 160 storey high rise tower, with a podium influenced by the deposition of marine sediments resulting
development around the base of the tower, including a 4-6 from a number of changes in sea level during relatively recent
storey garage. The client for the project is Emaar, a leading geological time. The country is generally relatively low-lying
developer based in Dubai. Once completed, the Burj Dubai (with the exception of the mountainous regions in the north-
Tower will be the world’s tallest building. It is founded on a east of the country), with near-surface geology dominated by
3.7m thick raft supported on bored piles, 1.5 m in diameter, deposits of Quaternary to late Pleistocene age, including
extending approximately 50m below the base of the raft. mobile Aeolian dune sands, evaporite deposits and marine
Figure 1 shows an artist’s impression of the completed tower. sands.
The site is generally level and site levels are related to Dubai
Municipality Datum (DMD). Dubai is situated towards the eastern edge of the geologically
stable Arabian Plate and separated from the unstable Iranian
The Architects and Structural Engineers for the project were Fold Belt to the north by the Arabian Gulf. The site is
Skidmore Owings and Merrill LLP (SOM) in Chicago. Hyder therefore considered to be located within a seismically active
Consulting (UK) Ltd (HCL) were appointed geotechnical area.
consultant for the works by Emaar and carried out the design
of the foundation system and an independent peer review has
been undertaken by Coffey Geosciences (Coffey). This paper GEOTECHNICAL INVESTIGATION & TESTING
describes the foundation design and verification processes, PROGRAM
and the results of the pile load testing programs. It also
compares the predicted settlements with those measured The geotechnical investigation was carried out in four phases
during construction. as follows:

Phase 1 (main investigation): 23 boreholes, in situ SPT’s, 40

pressuremeter tests in 3 boreholes, installation of 4 standpipe
piezometers, laboratory testing, specialist laboratory testing

Paper No. 1.47 1

and contamination testing – 1st June to 23rd July 2003; variations in the nature of the strata between boreholes.
Down-hole seismic testing was used to determine shear (S)
Phase 2 (main investigation): 3 geophysical boreholes with
wave velocities through the ground profile.
cross-hole and tomography geophysical surveys carried out
between 3 new boreholes and 1 existing borehole – 7th to 25th
August, 2003;
Phase 3: 6 boreholes, in situ SPT’s, 20 pressuremeter tests in
2 boreholes, installation of 2 standpipe piezometers and
laboratory testing – 16th September to 10th October 2003;
Phase 4: 1 borehole, in situ SPT’s, cross-hole geophysical
testing in 3 boreholes and down-hole geophysical testing in 1
borehole and laboratory testing.

The drilling was carried out using cable percussion techniques

with follow-on rotary drilling methods to depths between 30m
and 140m below ground level. The quality of core recovered
in some of the earlier boreholes was somewhat poorer than
that recovered in later boreholes, and therefore the defects
noted in the earlier rock cores may not have been
representative of the actual defects present in the rock mass.
Phase 4 of the investigation was targeted to assess the
difference in core quality and this indicated that the
differences were probably related to the drilling fluid used and
the overall quality of drilling.

Disturbed and undisturbed samples and split spoon samples

were obtained from the boreholes. Undisturbed samples were
obtained using double tube core barrels (with Coreliner) and
wire line core barrels producing core varying in diameter
between 57mm and 108.6mm.

Standard Penetration Tests (SPTs) were carried out at various

depths in the boreholes and were generally carried out in the
overburden soils, in weak rock or soil bands encountered in
the rock strata. Fig 1: Impression of Burj Dubai when Complete

Pressuremeter testing, using an OYO Elastmeter,

was carried out in 5 boreholes between depths of about 4m to Laboratory Testing
60m below ground level typically below the Tower footprint.
The geotechnical laboratory testing program consisted of two
The geophysical survey comprised cross-hole seismic survey, broad classes of test:
cross-hole tomography and down-hole geophysical survey.
The main purpose of the geophysical survey was to 1. Conventional tests, including moisture content,
complement the borehole data and provide a check on the Atterberg limits, particle size distribution, specific
results obtained from borehole drilling, in situ testing and gravity, unconfined compressive strength, point load
laboratory testing. index, direct shear tests, and carbonate content tests.
2. Sophisticated tests, including stress path triaxial,
The cross-hole seismic survey was used to assess compression resonant column, cyclic undrained triaxial, cyclic
(P) and shear (S) wave velocities through the ground profile. simple shear and constant normal stiffness (CNS)
Cross-hole tomography was used to develop a detailed direct shear tests. These tests were undertaken by a
distribution of P-wave velocity in the form of a vertical variety of commercial, research and university
seismic profile of P-wave with depth, and highlight any laboratories in the UK, Denmark and Australia.

Paper No. 1.47 2

GEOTECHNICAL CONDITIONS Groundwater levels are generally high across the site and
excavations were likely to encounter groundwater at
The ground conditions comprise a horizontally stratified approximately +0.0m DMD (approximately 2.5m below
subsurface profile which is complex and highly variable, due ground level).
to the nature of deposition and the prevalent hot arid climatic The ground conditions encountered in the investigation were
conditions. Medium dense to very loose granular silty sands consistent with the available geological information.
(Marine Deposits) are underlain by successions of very weak
to weak sandstone interbedded with very weakly cemented The ground profile and derived geotechnical design
sand, gypsiferous fine grained sandstone/siltstone and weak to parameters assessed from the investigation data are
moderately weak conglomerate/calcisiltite. summarized in Table 1.

Table 1. Summary of Geotechnical Profile and Parameters

Ult.
Level at top Thicknes Comp.
UCS Undrained Drained
Sub- Shaft
Strata Subsurface Material of stratum s Modulus* Modulus* Frictio
Strata
(m DMD) (m) (MPa) Eu (MPa) E’ (MPa) n fs
(kPa)
Medium dense silty
1a +2.50 1.50 - 34.5 30 -
Sand
1
Loose to very loose
1b +1.00 2.20 - 11.5 10 -
silty Sand

Very weak to
2 2 moderately weak -1.20 6.10 2.0 500 400 350
Calcarenite

Medium dense to
very dense Sand/ Silt
3a -7.30 6.20 - 50 40 250
with frequent
sandstone bands

Very weak to weak

3
3b Calcareous -13.50 7.50 1.0 250 200 250
Sandstone

Very weak to weak

3c Calcareous -21.00 3.00 1.0 140 110 250
Sandstone

Very weak to weak

gypsiferous
4 4 -24.00 4.50 2.0 140 110 250
Sandstone/
calcareous Sandstone

Paper No. 1.47 3

 Very weak to
moderately weak
5a Calcisiltite/ -28.50 21.50 1.3 310 250 285
Conglomeritic
Calcisiltite
5
Very weak to
moderately weak
5b Calcisiltite/ -50.00 18.50 1.7 405 325 325
Conglomeritic
Calcisiltite

Very weak to weak

6 6 Calcareous/ -68.50 22.50 2.5 560 450 400
Conglomerate strata

Weak to moderately
weak Claystone/
7 7 -91.00 >46.79 1.7 405 325 325
Siltstone interbedded
with gypsum layers
* Note that the Eu and E’ values relate to the relatively large strain levels in the strata.
E Value (MPa)

0.0 5000.0 10000.0 15000.0 20000.0 25000.0

0.00
Strata 1 (Sand)
Stiffness values from the pressuremeter reload cycle, the Strata 2 (Calcarenite)
Borehole 2 -
Pressuremeter
specialist tests and the geophysics are presented in Figure 2. Reload 1
Borehole 2 -
There is a fair correlation between the estimated stiffness -10.00
Strata 3a (Sand)
Pressuremeter
Reload 2

profiles from the pressuremeter and the specialist testing Borehole 3 -

Pressuremeter
Reload 1
Strata 3b (Sandstone)
results at small strain levels. -20.00
Borehole 3 -
Pressuremeter
Reload 2
Strata 3c (Sand)
Borehole 4 -
Pressuremeter
Non-linear stress-strain responses were derived for each strata Strata 4 (Gypsiferous Sandstone) Reload 1
Borehole 4 -
type using the results from the SPT’s, the pressuremeter, the -30.00
Pressuremeter
Reload 2

geophysics and the standard and specialist laboratory testing. Borehole 25 -

Pressuremeter
Reload 1
Strata 5a (Calcisiltite/ conglomeritic Calcisiltite)
Best estimate and maximum design curves were generated and -40.00 Borehole 25 -
Pressuremeter

the best estimate curves are presented in Figure 3. Reload 2

Borehole 28 -
Pressuremeter
Reload 1

-50.00 Borehole 28 -
An assessment of the potential for degradation of the stiffness Pressuremeter
Reload 2

of the strata under cyclic loading was carried out through a Strata 5b (Calcisiltite/ conglomeritic Calcisiltite)
Stress Path Txl at
0.01% strain

review of the CNS and cyclic triaxial specialist test results, -60.00 Stress Path Txl at
0.1% strain

and also using the computer program SHAKE91 (Idriss and Resonant Column
at 0.0001% strain
Sun, 1992) for potential degradation under earthquake loading. -70.00 Resonant Column
The results indicated that there was a potential for degradation at 0.001% strain

of the mass stiffness of the materials but limited potential for Resonant Column

Fig 2: Modulus Values vs Elevation

Strata 6 - (Calcareous/ Conglomeritic Strata) at 0.01% strain

degradation of the pile-soil interface. An allowance for -80.00

Geophysics E
Values (lowest)

degradation of the mass stiffness of the materials has been Adopted Small
Strain Design
incorporated in the derivation of the non-linear curves in -90.00
Values

Figure 3.

Paper No. 1.47 4

 Proposed Nonlinear Ground Strata Characteristics

4
FOUNDATION DESIGN
3.5
An assessment of the foundations for the structure was carried
3
out and it was clear that piled foundations would be
2.5
appropriate for both the Tower and Podium construction. An
2
initial assessment of the pile capacity was carried out using the
1.5
following design recommendations given by Horvath and
1 Kenney (1979), as presented by Burland and Mitchell (1989):
0.5

0
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14
Ultimate unit shaft resistance fs = 0.25 (qu) 0.5
Strain

Strata 2 Strata 3a Strata 3b Strata 3c & 4 Strata 5a, 5b, 6, 7

where fs is in kPa, and qu = uniaxial compressive strength in
MN/m2
Fig 3: Non-linear Stress-strain Curves
The adopted ultimate compressive unit shaft friction values for
the various site rock strata are tabulated in Table 1. The
ultimate unit pile skin friction of a pile loaded in tension was
taken as half the ultimate unit shaft resistance of a pile loaded
GEOTECHNICAL MODELS AND ANALYSES in compression.

A number of analyses were used to assess the response of the

foundation for the Burj Dubai Tower and Podium. The main The assessed pile capacities were provided to SOM and they
design model was developed using a Finite Element (FE) then supplied details on the layout, number and diameter of
program ABAQUS run by a specialist company KW Ltd, the piles. Tower piles were 1.5m diameter and 47.45m long
based in the UK. Other models were developed to validate with the tower raft founded at -7.55mDMD. The podium piles
and correlate the results from the ABAQUS model using were 0.9m diameter and 30m long with the podium raft being
software programs comprising REPUTE (Geocentrix, 2002), founded at -4.85mDMD. The thickness of the raft was 3.7m.
PIGLET (Randolph, 1996) and VDISP (OASYS Geo, 2001). Loading was provided by SOM and comprised 8 load cases
including four load cases for wind and three for seismic
The ABAQUS model comprised a detailed foundation mesh conditions.
of 500m by 500m by 90m deep. The complete model
The initial ABAQUS runs indicated that the strains in the
incorporated a ‘far field’ coarse mesh of 1500m by 1500m by
strata were within the initial small strain region of the non-
300m deep. A summary of the model set up is as follows:
linear stress strain curves developed for the materials. The
Soil Strata: Modeled as Von Mises material (pressure secant elastic modulus values at small strain levels were
independent), based on non-linear stress-strain curvesTower therefore adopted for the validation and sensitivity analyses
Piles: Modeled as beam elements connected to the soil strata carried out using PIGLET and REPUTE. A non-linear
analysis was carried out in VDISP using the non-linear stress
by pile-soil interaction elements. Class A load-settlement
predictions were used to calibrate the elements; strain curves developed for the materials.
Podium Piles: Beam elements fully bonded to the soil strata;
Tower and Podium Loadings: Applied as concentrated Linear and non-linear analyses were carried out to obtain
loadings at the column locations; predictions for the load distribution in the piles and for the
Tower raft submerged weight: Applied as a uniformly settlement of the raft and podium.
distributed load;
Tower Shearing Action: Applied as a body load to the tower The settlements from the FE analysis model and from VDISP
raft elements, in a direction to coincide with the appropriate have been converted from those for a flexible pile cap to those
wind action assumed; for a rigid pile cap for comparison with the REPUTE and
Building Stiffness Effect: Superstructure shear walls (not PIGLET models using the following general equation and are
shown in Table 2:
interrupted at door openings) were modeled as a series of
beam elements overlaid on the tower raft elements. The δrigid = 1/2 (δ centre + δ edge)flexible
moment of inertia was modified to simulate the stiffening
effect of the tower, as specified by SOM.

Paper No. 1.47 5

Table 2. Computed Settlements from Analyses opposite where the largest pile loads are concentrated towards
the edges of the pile group reducing towards the centre of the
Analysis Loadcase Settlement mm group. Similarly, the PIGLET and REPUTE standard pile
Method group analyses carried out indicated that the largest pile loads
Rigid Flexible are concentrated towards the edge of the pile cap.
FEA Tower 56 66
Only
(DL+LL)

REPUTE Tower 45 -
Only
(DL+LL)

PIGLET Tower 62 -
Only
(DL+LL)

VDISP Tower 46 72
Only
(DL+LL) Fig 4: Contours of Maximum Axial Load

The difference between the pile load distributions could be

Gratifyingly, the settlements from the FEA model correlated attributed to a number of reasons including:
acceptably well with the results obtained from REPUTE,
PIGLET and VDISP.
The FE, REPUTE and PIGLET models take account
of the pile-soil-pile interaction whereas SOM
A sensitivity analysis was carried out using the FE analysis modelled the soil as springs connected to the raft and
model and applying the maximum design soil strata non-linear piles using an S-Frame analysis.
stress-strain relationships. The results from the stiffer soil

The HCL FE analysis modelled the soil/rock using
strata response gave a 28% reduction in Tower settlement for
non-linear responses compared to the linear spring
the combined Dead load, live load and wind load case
stiffnesses assumed in the SOM analysis.
analyzed, from 85mm to 61mm.

The specified/assumed superstructure stiffening
The maximum and minimum pile loadings were obtained from effects on the foundation response were modelled
the FE analysis for all loading combinations. The maximum more accurately in the SOM analysis.
loads were at the corners of the three “wings” and were of the
order of 35 MN, while the minimum loads were within the
In reality the actual pile load distribution is expected to be
center of the group and were of the order of 12-13 MN. Figure
somewhere between the two models depending on the impact
4 shows contours of the computed maximum axial load. The
of the different modeling approaches.
impact of cyclic loading on the pile was an important
consideration and in order to address this, the load variation
above or below the dead load plus live load cases was
OVERALL STABILITY ASSESSMENT
determined. The maximum load variation was found to be
less than 10 MN.
The minimum centre-to-centre spacing of the piles for the
SOM carried out an analysis of the pile loads and a tower is 2.5 times the pile diameter. A check was therefore
comparison on the results indicates that although the carried out to ensure that the Tower foundation was stable
maximum pile loads are similar, the distribution is different. both vertically and laterally assuming that the foundation acts
The SOM calculations indicated that the largest pile loads are as a block comprising the piles and soil/rock. A factor of
in the central region of the Tower piled raft and decreasing safety of just less than 2 was assessed for vertical block
towards the edges. However, the FE analyses indicated the movement, excluding base resistance of the block while a

Paper No. 1.47 6

factor of safety of greater than 2 was determined for lateral area, and the piles were represented by a solid block
block movement excluding passive resistance. A factor of containing piles and soil. The axial stiffness of the
safety of approximately 5 was obtained against overturning of block was taken to be the same as that of the piles
the block. and the soil between them. The total dead plus live
loading was assumed to be uniformly distributed. The
soil layers were assumed to be Mohr Coulomb
LIQUEFACTION ASSESSMENT materials, with the modulus values as shown in Table
3, and values of cohesion taken as 0.5 times the
An assessment of the potential for liquefaction during a estimated unconfined compressive strength. The
seismic event at the Burj Dubai site has been carried out using main purpose of this analysis was to calibrate and
the Japanese Road Association Method and the method of check the second, and more detailed, analysis, using
Seed et al (1984). Both approaches gave similar results and the computer program for pile group analysis, PIGS
indicated that the Marine Deposits and sand to 3.5m below (Poulos, 2002).
ground level (from +2.5 m DMD to –1.0 m DMD) could
2. An analysis using PIGS was carried out for the tower
potentially liquefy. However the foundations of the Podium
alone, to check the settlement with that obtained by
and Tower structures were below this level. Consideration
FLAC. In this analysis, the piles were modeled
was however required in the design and location of buried
individually, and it was assumed that each pile was
services and shallow foundations which were within the top
subjected to its nominal working load of 30MN. The
3.5m of the ground. Occasional layers within the sandstone
stiffness of each pile was computed via the program
layer between –7.3 m DMD and –11.75 m DMD could
DEFPIG (Poulos, 1990), allowing for contact
potentially liquefy. However, taking into account the imposed
between the raft section above the pile and the
confining stresses at the foundation level of the Tower this
underlying soil. The pile stiffness values were
was considered to have a negligible effect on the design of the
assumed to vary hyperbolically with increasing load
Tower foundations. The assessed reduction factor to be
level, using a hyperbolic factor (Rf) of 0.4.
applied to the soil strength parameters, in most cases, was
found to be equal to 1.0 and hence liquefaction would have a 3. Finally, an analysis of the complete tower-podium
minimal effect upon the design of the Podium foundations. foundation system was cried out using the program
However, consideration was given in design for potential PIGS, and considering all 926 piles in the system.
downdrag loads on pile foundations constructed through the Again, each of the piles was subjected to its nominal
liquefiable strata. working load.

INDEPENDENT VERIFICATION ANALYSES

FLAC & PIGS Results For The Tower Alone
The geotechnical model used in the verification analyses is
summarized in Table 3. The parameters were assessed
Because of the difference in shape between the actual
independently on the basis of the available information and
foundation and the equivalent circular foundation, only the
experience gained from the nearby Emirates project (Poulos
maximum was considered for comparison purposes. The
and Davids, 2005). In general, this model was rather more
following results were obtained for the central settlement:
conservative than the original model employed for the design.
In particular, the ultimate end bearing capacity was reduced
FLAC analysis, using an equivalent block to represent the
together with the Young’s modulus in several of the upper
piles: 72.9mm
layers, and the presence was assumed of a stiffer layer, with a
modulus of 1200 MPa below RL –70m DMD, to allow for the
PIGS analysis, modeling all 196 piles: 74.3mm
fact that the strain levels in the ground decrease with
increasing depth.
Thus, despite the quite different approaches adopted, the
computed settlements were in remarkably good agreement. It
The following three-stage approach was employed for the
should be noted that the computed settlement is influenced by
independent verification process:
the assumptions made regarding the ground properties below
1. The commercially available computer program
the pile tips. For example, if in the PIGS analysis the modulus
FLAC was used to carry out an axisymmetric
of the ground below RL-70m DMD was taken as 400 MPa
analysis of the foundation system for the tower. The
foundation plan was represented by a circle of equal

Paper No. 1.47 7

(rather than 1200 MPa), the computed settlement at the centre Figures 6 shows the settlement profile across a section through
of the tower would increase to about 96 mm. the centre of the tower. The notable feature of this figure is
that the settlements reduce rapidly outside the tower area, and
become of the order of 10-12 mm for much of the podium
PIGS Results For Tower & Podium area.

Figure 5 shows the contours of computed settlement for the The settlements of the tower computed from the independent
entire area. It can be seen that the maximum settlements are verification process agreed reasonably well with those
concentrated in the central area of the tower. obtained for the original design, as reported above.

Table 3. Summary of Geotechnical Model for Independent Verification Analyses

Stratum Description RL Undrained Drained Ultimate Skin Ultimate End

Number Range Modulus Modulus E’ Friction Bearing
DMD Eu MPa MPa kPa MPa
1a Med. Dense silty sand +2.5 to 30 25 - -
+1.0

1b Loose-v.loose silty sand _1.0 to 12.5 10 - -

–1.2

2 Weak-mod. Weak -1.2 to 400 325 400 4.0

calcarenite –7.3

3 V. weak calc. -7.3 to 190 150 300 3.0

Sandstone -24

4 V. weak-weak -24 to 220 175 360 3.6

sandstone/calc. –28.5
Sandstone
5A V. weak-weak-mod. -28.5 to 250 200 250 2.5
Weak –50
calcisiltite/conglom.
5B V. weak-weak-mod. -50 to – 275 225 275 2.75
Weak 70
calcisiltite/conglom
6 Calcareous siltstone -70 & 500 400 375 3.75
below

Paper No. 1.47 8

 Fig 5: Computed Settlement Contours for Tower and Podium

Fig 6: Computed Settlement Across Section Through Centre of Tower

Paper No. 1.47 9

CYCLIC LOADING EFFECTS the piles, and that the effects of cyclic loading on both
capacity and settlement were unlikely to be significant.
The possible effects of cyclic loading were investigated via the
following means:
• Cyclic triaxial laboratory tests; PILE LOAD TESTING
• Cyclic direct shear tests;
• Cyclic Constant Normal Stiffness (CNS) laboratory Two programs of static load testing were undertaken for the
tests; Burj Dubai project:
• Via an independent theoretical analysis carried out by • Static load tests on seven trial piles prior to
the independent verifier. foundation construction.
• Static load tests on eight works piles, carried out
The cyclic triaxial tests indicated that there is some potential during the foundation construction phase (i.e. on
for degradation of stiffness and accumulation of excess pore about 1% of the total number of piles constructed).
pressure, while the direct shear tests have indicated a In addition, dynamic pile testing was carried out on 10 of the
reduction in residual shear strength, although these were works piles for the tower and 31 piles for the podium, i.e. on
carried out using large strain levels which are not about 5% of the total works piles. Sonic integrity testing was
representative of the likely field conditions. also carried out on a number of the works piles. Attention here
is focused on the static load tests.
The CNS tests indicated that there is not a significant potential
for cyclic degradation of skin friction, provided that the cyclic
shear stress remains within the anticipated range. Preliminary Pile Testing Program

The independent analysis of cyclic loading effects was The details of the piles tested within this program are
undertaken using the approach described by Poulos (1988), summarized in Table 4. The main purpose of the tests was to
and implemented via the computer program SCARP (Static assess the general load-settlement behaviour of piles of the
and Cyclic Axial Response of Piles). This analysis involved a anticipated length below the tower, and to verify the design
number of simplifying assumptions, together with parameters assumptions. Each of the test piles was different, allowing
that were not easily measured or estimated from available various factors to be investigated, as follows:
data. As a consequence, the analysis was indicative only. • The effects of increasing the pile shaft length;
Since the analysis of the entire foundation system was not • The effects of shaft grouting;
feasible with SCARP, only a typical pile (assumed to be a • The effects of reducing the shaft diameter;
single isolated pile) with a diameter of 1.5m and a length of • The effects of uplift (tension) loading;
48m was considered. The results were used to explore the • The effects of lateral loading;
relative effects of the cyclic loading, with respect to the case • The effect of cyclic loading.
of static loading. The piles were constructed using polymer drilling fluid, rather
than the more conventional bentonite drilling fluid. As will be
It was found that a loss of capacity would be experienced shown below, the use of the polymer appears to have led to
when the cyclic load exceeded about ± 10MN. The maximum piles whose performance exceeded expectations.
loss of capacity (due to degradation of the skin friction) was of
the order of 15-20%. The capacity loss was relatively Strain gauges were installed along each of the piles, enabling
insensitive to the mean load level, except when the mean load detailed evaluation of the load transfer along the pile shaft,
exceeded about 30 MN. It was predicted that, at a mean load and the assessment of the distribution of mobilized skin
equal to the working load and under a cyclic load of about friction with depth along the shaft. The reaction system
25% of the working load, the relative increase in settlement provided for the axial load tests consisted of four or six
for 10 cycles of load would be about 27%. adjacent reaction piles (depending on the pile tested), and
these reaction piles had the potential to influence the results of
The indicative pile forces calculated from the ABAQUS finite the pile load tests via interaction with the test pile through the
element analysis of the structure suggested that cyclic loading soil. The possible consequences of this are discussed
of the Burj Tower foundation would not exceed ± 10MN. subsequently.
Thus, it seemed reasonable to assume that the effects of cyclic
loading would not significantly degrade the axial capacity of

Paper No. 1.47 10

 The skin friction values down to about RL-30m DMD appear
Table 4. Summary of Pile Load Tests – Preliminary Pile to be ultimate values, i.e. the available skin friction has been
Testing fully mobilized.
The skin friction values below about RL-30m DMD do not
Pile Pile Pile Side Test Type appear to have been fully mobilized, and thus were assessed to
No. Diam. Length Grouted be below the ultimate values.
m m ?
The original assumptions appear to be comfortably
TP1 1.5 45.15 No Compression conservative within the upper part of the ground profile.
TP2 1.5 55.15 No Compression Shaft grouting appeared to enhance the skin friction developed
along the pile.
TP3 1.5 35.15 Yes Compression
TP4 0.9 47.10 No Compression Because the skin friction in the lower part of the ground
(cyclic) profile does not appear to have been fully mobilized, it was
recommended that the original values (termed the “theoretical
TP5 0.9 47.05 Yes Compression ultimate unit skin friction”) be used in the lower strata. It was
also recommended that the “theoretical” values in the top
TP6 0.9 36.51 No Tension
layers (Strata 2 and 3a) be used because of the presence of the
TP7A 0.9 37.51 No Lateral casing in the tests would probably have given skin friction
values that may have been too low. For Strata 3b, 3c and 4, the
minimum measured skin friction values were used for the final
design.
Ultimate Axial Load Capacity
Maximum Skin Friction Values (kPa)
0 100 200 300 400 500 600 700 800 900
0

None of the 6 axial pile load tests appears to have reached its
ultimate axial capacity, at least with respect to geotechnical -10

resistance. The 1.5m diameter piles (TP1, TP2 and TP3) were -20

loaded to twice the working load, while the 0.9m diameter test
Elevation (m DMD)

piles TP4 and TP6 were loaded to 3.5 times the working load, -30

and TP5 was loaded to 4 times working load. With the -40

exception of TP5, none of the other piles showed any strong

-50

indication of imminent geotechnical failure. Pile TP5 showed

a rapid increase in settlement at the maximum load, but this -60

TP1 TP2 TP4 Theoretical From PPT Program Strata Levels

was attributed to structural failure of the pile itself. From a

design viewpoint, the significant finding was that, at the Fig 7: Measured and Design Values of Shaft Friction
working load, the factor of safety against geotechnical failure
appeared to be in excess of 3, thus giving a comfortable
margin of safety against failure, especially as the raft would
also provide additional resistance to supplement that of the Ultimate End Bearing Capacity
piles.
None of the load tests was able to mobilize any
significant end bearing resistance, because the skin
Ultimate Shaft Friction
friction appeared to be more than adequate to resist loads
From the strain gauge readings along the test piles, the
well in excess of the working load. Therefore, no
mobilized skin friction distribution along each pile was conclusions could be reached about the accuracy of the
evaluated. Figure 7 summarizes the ranges of skin friction estimated end bearing component of pile capacity. For
deduced from the measurements, together with the original the final design, the length of the piles was increased
design assumptions and the modified design recommendations where the proposed pile toe levels were close to or
made after the preliminary test results were evaluated. The within the gypsiferous sandstone layer (Stratum 4).
following comments can be made:
This was the case for the 0.9m diameter podium piles. It was

Paper No. 1.47 11

considered prudent to have the pile toes founded below this engineering properties of this layer (e.g. via solution of the
stratum, to allow for any potential long-term degradation of gypsum) that could reduce the capacity of the piles.

Table 5. Summary of Pile Load Test Results – Axial Loading

Pile Number Working Load Max. Load MN Settlement at W. Settlement at Stiffness at W. Stiffness at Max.
MN Load mm Max. Load mm Load MN/m Load MN/m
TP1 30.13 60.26 7.89 21.26 3819 2834
TP2 30.13 60.26 5.55 16.85 5429 3576
TP3 30.13 60.26 5.78 20.24 5213 2977
TP4 10.1 35.07 4.47 26.62 2260 1317
TP5 10.1 40.16 3.64 27.45 2775 1463
TP6 -1.0 -3.5 -0.65 -4.88 1536 717

high stiffness of the pile may not reflect the true stiffness of
Load-Settlement Behaviour the pile beneath the structure. The mechanisms of such
interaction are discussed by Poulos (2000).
Table 5 summarises the measured pile settlements at the
working load and at the maximum test load, and the
corresponding values of pile head stiffness (load/settlement). Pile Axial Stiffness Predictions
The following observations are made:
• The measured stiffness values are relatively large, “Class A” predictions of the anticipated load-settlement
and are considerably in excess of those anticipated. behaviour were made prior to the construction of the
• As expected, the stiffness is greater for the larger preliminary test piles. The designer used the finite element
diameter piles. program ABAQUS, while the independent verifier used the
• The stiffness of the shaft grouted piles (TP3 and computer program PIES (Poulos, 1989). No allowance was
TP5) is greater than that of the corresponding made for the effects of interaction from the reaction piles.
ungrouted piles. There was close agreement between the predicted curves for
the 1.5m diameter piles extending to RL-50m, but for the 0.9m
diameter piles extending to RL-40m, the agreement was less
close, with the designer predicting a somewhat softer
Effect of Reaction Piles behaviour than the independent verifier.

On the basis of the experience gained in the nearby Emirates The measured load-settlement behaviour was considerably
Project (Poulos and Davids, 2005) site), it had been expected stiffer than either of the predictions. This is shown in Figure 8,
that the pile head stiffness values for the Burj Dubai piles which compares the measured stiffness values with the
would be somewhat less than those for the Emirates Towers, predicted values, at the working load. As mentioned above,
in view of the apparently inferior quality of rock at the Burj the high measured stiffness may be, at least partly, a
Dubai site. This expectation was certainly not realized, and it consequence of the effects of the adjacent reaction piles. An
is possible that the improved performance of the piles in the analysis of the effects of these reaction piles on the settlement
present project may be attributable, at least in part, to the use of pile TP1 revealed that the presence of the reaction piles
of polymer drilling fluid, rather than bentonite, in the could reduce the settlement at the working load of 30MN by
construction process. However, it was also possible that at 30%. In other words, the real stiffness of the piles might be
least part of the reason for the high stiffness values is related only about 70% of the values measured from the load test.
to the interaction effects of the reaction piles. When applying a This would then reduce the stiffness to a value which is more
compressive load to the test pile, the reaction piles will in line with the stiffness values experienced in the Emirates
experience a tension and a consequent uplift, which will tend project, where the reaction was provided by a series of
to reduce the settlement of the test pile. Thus, the apparent

Paper No. 1.47 12

inclined anchors that would have had a very small degree of Table 6. Summary of Displacement Accumulation for Cyclic
interaction with the test piles. Loading
6000
Pile Mean Cyclic No. of SN/S1
5000
Number Load/Pw Load/Pw Cycles
(N)
4000
TP1 1.0 ±0.5 6 1.12
Test Pile Num ber

At Working Load
3000 At Maximum Load
Calc. At Working Load
TP2 1.0 ±0.5 6 1.25
2000
TP3 1.0 ±0.5 6 1.25
1000
TP4 1.25 ±0.25 9 1.25
0
TP1 TP2 TP3 TP4 TP5 TP6
TP5 1.25 ±0.25 6 1.3
Stiffness MN/m

TP6 1.0 ±0.5 6 1.1

Fig 8: Measured and Predicted Pile Head Stiffness Values
Note: Pw = working load; SN = settlement after N cycles;
S1=settlement after 1 cycle
Uplift versus Compression Loading

On the basis of the tension test on pile TP6, the ultimate skin Lateral Loading
friction in tension was taken as 0.5 times that for compression.
It is customary to allow for a reduction in skin friction for One lateral load test was carried out, on pile TP7A, with the
piles in granular soils or rocks subjected to uplift. De Nicola pile being loaded to twice the working load (50t). At the
and Randolph (1993) have developed a theoretical relationship working lateral load of 25t, the lateral deflection was about
between the tensile and compressive skin friction values, and 0.47mm, giving a lateral stiffness of about 530 MN/m, a value
have shown that this relationship depends on the Poisson’s which was consistent with the designer’s predictions using the
ratio of the pile, the relative stiffness of the pile to the soil, the program ALP (Oasys, 2001). An analysis of lateral deflection
interface friction characteristics and the pile length to diameter was also carried out by the independent verifier using the
ratio. This theoretical relationship was applied to the Burj program DEFPIG. In this latter analysis, the Young’s modulus
Dubai case, and the calculated ratio of tension to compression values for lateral loading were assumed to be 30% less than
skin friction was about 0.6, which was reasonably consistent the values for axial loading, while the ultimate lateral pile-soil
with the assumption of 0.5 made in the design. pressure was assumed to be similar to the end bearing capacity
of the pile, with allowances being made for near-surface
effects. These calculations indicated a lateral movement of
Cyclic Loading Effects about 0.7mm at 25t load, which is larger than the measured
deflection, but of a similar order.
In all of the axial load tests, a relatively small number of
cycles of loading was applied to the pile after the working load Thus, pile TP7A appeared to perform better than anticipated
was reached. Table 6 summarizes the test results inferred from under the action of lateral loading, mirroring the better-than-
the load-settlement data. The settlement after cycling was expected performance of the test piles under axial load.
related to the settlement for the first cycle, both settlements However, there may again have been some effect of the
being at the maximum load of the cycling process. It can be reaction system used for the test, as the reaction block will
seen that there is an accumulation of settlements under the develop a surface shear which will tend to oppose the lateral
action of the cyclic loading, but that this accumulation is deflection of the test pile.
relatively modest, given the relatively high levels of mean and
cyclic stress that have been applied to the pile (in all cases, the
maximum load reached is 1.5 times the working load). Works Pile Testing Program

These results are consistent with the assessments made during A total of eight works pile tests were carried, including two
design that cyclic loading effects would be unlikely to be 1.5m diameter piles and six 0.9m diameter piles. All pile tests
significant for this building. were carried out in compression, and each pile was tested

Paper No. 1.47 13

approximately 4 weeks after construction. The piles were to be in excess of the predicted values, although none of the
tested to a maximum load of 1.5 times the working load. tests fully mobilized the available geotechnical resistance. The
works piles performed even better than the preliminary trial
The following observations were made from the test results: piles, and demonstrated almost linear load-settlement
• The pile head stiffness of the works piles was behaviour up to the maximum test load of 1.5 times working
generally larger than for the trial piles. load.
• None of the works piles reached failure, and indeed,
the load-settlement behaviour up to 1.5 times the Shaft grouting appeared to have enhanced the load-settlement
working load was essentially linear, as evident from response of the piles, but it was assessed that shaft grouting
the relatively small difference in stiffness between would not need to be carried out for this project, given the
the stiffness values at the working load and 1.5 times very good performance of the ungrouted piles.
the working load. In contrast, the relative difference
between the two stiffnesses was considerably greater The inferences from the pile load test data are that the design
for the preliminary trial piles. estimates of capacity and settlement may be conservative,
although it must be borne in mind that the overall settlement
At least three possible explanations could be offered for the behaviour (and perhaps the overall load capacity) are
greater stiffness and improved load-settlement performance of dependent not only on the individual pile characteristics, but
the trial piles: also on the characteristics of the ground within the zone of
1. The level of the bottom of the casing was higher influence of the structure.
for the works piles than for the trial piles (about
3.5-3.6 m higher), thus leading to a higher skin
SETTLEMENT PERFORMANCE DURING
friction along the upper portion of the shaft;
CONSTRUCTION
2. A longer period between the end of construction
and testing of the works piles (about 4 weeks, The settlement of the Tower raft has been monitored since
versus about 3 weeks for the trial piles); completion of concreting. The stress conditions within the raft
3. Natural variability of the strata. have been determined with the placement of strain rosettes at
the top and base of the raft. In addition three pressure cells
have been placed at the base of the raft and five piles have
Cyclic loading was undertaken on two of the works piles, and
been strain gauged to determine the load distribution between
it was observed that there was a relatively small amount of
and down the pile. This paper presents only the current
settlement accumulation due to the cyclic loading, and
situation on the settlement. The results from the strain gauges
certainly less than that observed on TP1 or the other trial piles
will be presented at a later date.
(see Table 6). The smaller amount of settlement accumulation
could be attributed to the lower levels of mean and cyclic
A summary of the settlements to 18 March 2007 is shown on
loading applied to the works piles (which were considered to
Figure 9 which also shows the final predicted settlement
be more representative of the design condition) and also to the
profile from the design. At that date, it is estimated that about
greater capacity that the works piles seem to possess. Thus,
75% of the dead load would have been acting on the
the results of these tests reinforced the previous indications
foundation. It should be noted that the monitored figures do
that the cyclic degradation of capacity and stiffness at the pile
not include the impact of the raft, cladding and live loading
– soil interface appeared to be negligible.
which will total in excess of 20% of the overall mass. It will
be seen that the measured settlements are considerably less
than those predicted during the design process, although there
Summary
remains some dead and live load to be applied to the
foundation system.
Both the preliminary test piling program and the tests on the
works piles provided very positive and encouraging
information on the capacity and stiffness of the piles. The
CONCLUSIONS
measured pile head stiffness values were well in excess of
those predicted. The interaction effects between the test piles
This paper has outlined the processes followed in the design of
and the reaction piles may have contributed to the higher
the foundations for the Burj Dubai and the independent
apparent pile head stiffnesses, but the piles nevertheless
verification of the design. The ground conditions at the site
exceeded expectations. The capacity of the piles also appeared
comprise a horizontally stratified subsurface profile which is

Paper No. 1.47 14

complex whose properties are highly variable with depth. A stiffnesses. The capacity of the piles also appears to be in
piled raft foundation system, with the piles socketed into weak excess of that predicted, and none of the tests appears to have
rock, has been employed and the design of the foundation is fully mobilized the available geotechnical resistance.
found to be governed primarily by the tolerable settlement of
the foundation rather than the overall allowable bearing The works piles have performed even better than the
capacity of the foundation. The capacity of the piles will be preliminary trial piles, and have demonstrated almost linear
derived mainly from the skin friction developed between the load-settlement behaviour up to the maximum test load of 1.5
pile concrete and rock, although limited end bearing capacity times working load.
will be provided by the very weak to weak rock at depth.
The settlements measured during construction are consistent
The estimated maximum settlement of the tower foundation, with, but smaller than, those predicted, and overall, the
calculated using the various analysis tools are in reasonable performance of the piled raft foundation system has exceeded
agreement, with predicted settlements of the tower ranging expectations to date.
from 45mm to 62mm. These results are considered to be
within an acceptable range.
Acknowledgements
The maximum settlement predicted by ABAQUS for the
tower and podium foundation compares reasonably well with The Authors would like to thank Mr Kamiran Ibrahim, Ms
the maximum settlement estimated by the revised PIGS Catherine Murrells and Ms Louise Baker from Hyder, and
analysis carried out during the independent verification Frances Badelow and Muliadi Merry from Coffey, for their
process. invaluable contribution to the design and review of the
foundation for the Burj Tower. The authors would also like to
There is a potential for a reduction in axial load capacity and thank Mr Bill Baker, Mr Stan Korista and Mr Larry Novak of
stiffness of the foundation strata under cyclic loading; but SOM for their inputs and useful discussions through the
based on the pile load test data, laboratory tests and on design period.
theoretical analyses, it would appear that the cyclic
degradation effects at the pile-soil interface are relatively
small. REFERENCES

Settlement in Wing C Burland, J. B., & Mitchell, J. M. (1989). Piling and Deep
Distance along wing cross-section (m) Foundations. Proc. Int. Conf. on Piling and Deep Foundations,
0
0 10 20 30 40 50 60 70 80
London, May 1989.
-10

-20 27-Jun-06
16-Jul-06
16-Aug-06
De Nicola, A. and Randolph, M.F. (1993). “Tensile and
-30 18-Sep-06
16-Oct-06
compressive shaft capacity of piles in sand”. Jnl. Geot. Eng.,
Settlement (mm)

14-Nov-06
-40
19-Dec-06
16-Jan-07
ASCE, Vol.119 (12): 1952-1973.
-50 19-Feb-07
18-Mar-07
Design
-60
Fleming, W. G. K., Weltman, A. J., Randolph, M.F. , Elson,
-70 W. K. (1994). Piling Engineering.
-80

-90 Geocentrix Ltd. (2002). “Repute Version 1 Reference

Fig 9: Measured and Computed Settlements for Wing C. Manual”.

Both the preliminary test piling program and the tests on the Horvath, R. and Kenney, T.C. (1979). “Shaft resistance of
works piles have provided very positive and encouraging rock-socketed drilled piers”. Presented at ASCE Annual
information on the capacity and stiffness of the piles. Convention, Atlanta, GA, preprint No. 3698.

The measured pile head stiffness values have been well in Idriss, I.M, Sun, J.I. (1992). “User’s Manual for SHAKE91”,
excess of those predicted, and those expected on the basis of Structures Division, Building and Fire Research Laboratory,
the experience with the nearby Emirates Towers. However, National Institute of Standards and Technology, Gaithersburg,
the interaction effects between the test piles and the reaction Maryland and Center for Geotechnical Modeling, Department
piles may have contributed to the higher apparent pile head of Civil & Environmental Engineering, University of
California, Davis, California.

Paper No. 1.47 15

OASYS Geo (2001), “ALP 17 GEO Suite for Windows”.

OASYS Geo (2001), “VDISP 17 GEO Suite for Windows”.

Poulos, H.G. (1988). “Cyclic Stability Diagram for Axially

Loaded Piles”. Jnl. Geot. Eng., ASCE, Vol. 114 (8): 877-895.

Poulos, H.G. (1989). “PIES User’s Manual”. Centre for

Geotechnical Research, University of Sydney, Australia.

Poulos, H.G. (1990). “DEFPIG Users Manual”. Centre for

Geotechnical Research, University of Sydney, Australia.

Poulos, H.G. (2000). “Pile testing – from the designer’s

viewpoint”. STATNAMIC Loading Test ’98, Kusakabe,
Kuwabara & Matsumoto (eds), Balkema, Rotterdam, 3-21.

Poulos, H.G. (2002). “Prediction of Behaviour of Building

Foundations due to Tunnelling Operations”. Proc. 3rd Int.
Symp. On Geot. Aspects of Tunnelling in Soft Ground,
Toulouse, Preprint Volume, 4.55-4.61.

Poulos, H.G. and Davids, A.J. (2005). “Foundation Design for

the Emirates Twin Towers, Dubai”. Can. Geot. Jnl., 42: 716-
730.

Randolph, M.F. (1996), “PIGLET Analysis and Design of Pile

Groups”. The University of Western Australia
Seed, H.B., Tokimatsu, K., Harder, L.F. and Chung, R.M.
(1984). “The influence of SPT procedure in soil liquefaction
resistance evaluation”. EERC-84/15, Univ. of California,
Berkeley.

Zhang, L. and Einstein, H. (1998). End bearing capacity of

drilled shafts in rock. Jnl. Geot. Eng., ASCE, 124(7): 574-
584.

Paper No. 1.47 16