7 Case study 7: Burj Khalifa piled raft
7.1 General
Burj Khalifa is a 163storey skyscraper in Dubai, United Arab Emirates. The total height of the building is 829.8 [m], with a podium development at its base, including a 4 to 6storey garage. With a total height of 829.8 [m] and a roof height (excluding antenna) of 828 [m], Burj Khalifa has been the tallest structure and building in the world since its topping out in late 2008, Figure 71. Burj Khalifa is located on a 42 000 [m^{2}] site. The tower is founded on a 3.7 [m] thick raft supported on 192 bored piles, 1.5 [m] in diameter, extending 47.45 [m] below the base of the raft; podium structures are founded on a 0.65 [m] thick raft (increased to 1 [m] at column locations) supported on 750 bored piles, 0.9 [m] in diameter, extending 30–35 [m] below the base of the raft. The tower raft consists of three wings each is 50 [m] long and 25 [m] wide forming an area of 3305 [m^{2}]. Figure 72 shows an isometric view of Burj Khalifa Tower foundation system and a plan for pile locations.
Extensive studies using different calculation methods were carried out by Poulos and Bunce (2008), Badelow & Poulos (2016) and Russo etc. al. (2013).


Figure 72 Burj Khalifa Tower Foundation system
7.2 Analysis of the piled raft
Using the available data and results of the Burj Khalifa piled raft, which have been discussed in detail in the previous references, the nonlinear analyses of piled raft in ELPLA are evaluated and verified using the following loadsettlement relations of piles, El Gendy et al. (2006) and El Gendy (2007):
1 Hyperbolic Function for LoadSettlement Curve.
2 Given LoadSettlement Curve.
The foundation system is analyzed as an elastic piled raft in which the raft is considered as an elastic plate supported on equal rigid piles.
A series of comparisons are carried out to evaluate the nonlinear analyses of piled raft for loadsettlement relations of piles. In which, results of other analytical solutions and measurements are compared with those obtained by ELPLA.
7.3 FENet
The raft is divided into triangular elements with a maximum length of 2.0 [m] as shown in Figure 73. Piles are divided into five elements with 9.49 [m] length.
7.4 Loads
Only longterm conditions have been considered, and for most of the early analyses, an average load per pile of 23.21 [MN] has been used (this is a representative of the design dead and live loads) and has been applied as an uniformly distributed load on the tower raft of about 1250 [kPa].
7.5 Pile and raft material
The raft is 3.7 [m] thick and was poured utilizing C50 (cube strength) selfconsolidating concrete. The Tower raft is supported by 192 bored castinplace piles. The C60 selfconsolidating concrete piles are 1.5 [m] in diameter and 47.45 [m] long.
The following values were used as pile and raft material:
For the raft:
Modulus of elasticity E_{p} = 33234 [MN/m^{2}]
Poisson's ratio v_{p} = 0.167 []
Unit weight γ_{b} = 23.60 [kN/m^{3}]
For piles:
Modulus of elasticity E_{p} = 36406 [MN/m^{2}]
Unit weight γ_{b} = 23.60 [kN/m^{3}]
7.6 Soil properties
The ground conditions comprise a horizontally stratified subsurface profile which is complex and highly variable, due to the nature of deposition and the prevalent hot arid climatic conditions. Medium dense to very loose granular silty sands (Marine Deposits) are underlain by successions of very weak to weak sandstone interbedded with very weakly cemented sand, gypsiferous fine grained sandstone/siltstone and weak to moderately weak conglomerate/calcisiltite.
Groundwater levels are generally high across the site and excavations were likely to encounter groundwater at approximately 2.5 [m] below ground level.
The drilling was carried out using cable percussion techniques with followon rotary drilling methods to depths between 30 [m] and 140 [m] below ground level.
The ground profile and derived geotechnical design parameters assessed from the investigation data are summarized in Table 71.
Table 71 Summary of Geotechnical Profile and Parameters
Strata 
SubStrata 
Subsurface Material 
Level at top of stratum
[m DMD] 
Thickness
H [m] 
UCS
q_{s} [MPa] 
Undrained Modulus
E_{u} [MPa] 
Ult. Comp. Shaft Frict. f_{s} [kPa] 
1 
1a 
Medium dense silty Sand 
+2.50 
1.50 
 
34.5 
 
1b 
Loose to very loose silty Sand 
+1.00 
2.20 
 
11.5 
 

2 
2 
Very weak to moderately weak Calcarenite 
1.20 
6.10 
2.0 
500 
350 
3 
3a 
Medium dense to very dense Sand/ Silt with frequent sandstone bands 
7.30 
6.20 
 
50 
250 
3b 
Very weak to weak Calcareous Sandstone 
13.50 
7.50 
1.0 
250 
250 

3c 
Very weak to weak Calcareous Sandstone 
21.00 
3.00 
1.0 
140 
250 

4 
4 
Very weak to weak gypsiferous Sandstone/ calcareous Sandstone 
24.00 
4.50 
2.0 
140 
250 
5 
5a 
Very weak to moderately weak Calcisiltite/ Conglomeritic Calcisiltite 
28.50 
21.50 
1.30 
310 
285 
5b 
Very weak to moderately weak Calcisiltite/ Conglomeritic Calcisiltite 
50.00 
18.50 
1.70 
405 
325 

6 
6 
Very weak to weak Calcareous/ Conglomerate strata 
68.50 
22.50 
2.50 
560 
400 
7 
7 
Weak to moderately weak Claystone/ Siltstone 
91.00 
>46.79 
1.70 
405 
325 
To carry out the analysis, the subsoil under the raft is considered as indicated in the boring log of Figure 74 that consists of 12 soil layers. The total depth under the ground surface is taken to be 140 [m].
7.7 Results
As examples for results of different analyses by ELPLA, Figure 78 and Figure 77 show the settlement for elastic piled raft of Burj Khalifa using methods: "Hyperbolic Function for LoadSettlement Curve" and "Given LoadSettlement Curve from pileload test", respectively. Besides, Figure 79, Figure 710 and Figure 711 show selfsettlement S_{v}, interaction settlement S_{rv} and total settlement S_{r }of piles using the method "Given LoadSettlement Curve from pileload test".
7.8 Measurements and other results
7.8.1 Measured settlement
The construction of Burj Khalifa began on 6 January 2004, with the exterior of the structure completed on 1 October 2009. According to Badelow & Poulos (2016) the settlement of the tower raft was monitored from completion of concreting till 18 February 2008. The recorded maximum settlement at 18 February 2008 was 43 [mm] under nearly 80 % of the building load.
A comparison is presented between the measured settlement on 18 February 2008 under 80% of the total load and that computed by ELPLA using Method: "Given LoadSettlement Curve". Figure 712 shows a comparison between measured settlement (Feb. 2008) and computed settlement under 80 % of the total load at a cross section of the Wing c, while 0 shows a comparison between extreme values of measured settlement and that calculated for the same case.
Table 72 Comparison between measured settlement at February 2008 and that calculated by ELPLA under 80 % of the total load
Method 
S_{max.} [mm] 
S_{min.} [mm] 
S_{Diff}_{.} [mm] 
Measured (18 February 2008) 
43 
29 
14 
ELPLA – Method: "Given LoadSettlement Curve" 
48 
24 
24 
Figure 713 shows contours of measured settlement [mm] at February 2008 and that calculated by ELPLA under 80 % of the total load using method "Given LoadSettlement Curve"


The above comparison of the piled raft under 80 % of the total load illustrates that the maximum and minimum results of ELPLA are in good agreement with the measured settlement with difference not exceed 1 [cm]. The measured differential settlement is considerably smaller than that computed because the building stiffness is not considered in ELPLA analysis in this case, which would reduce the differential settlement.
7.8.2 Calculated final settlement
Several analyses were used to assess the response of the foundation for the Burj Khalifa Tower and Podium. The main design model was developed using a Finite Element (FE) program ABAQUS run by a specialist company KW Ltd, based in the UK. Other models were developed to validate and correlate the results from the ABAQUS model using other software programs. The design values of settlement were presented by Poulos and Bunce (2008).
Russo etc. al. (2013) deals with the reassessment of foundation settlements for the Burj Khalifa Tower in Dubai. Reassessment was carried out using the computer program Nonlinear Analysis of Piled Rafts NAPRA with neglecting the structure stiffness effect on raft settlement.
A comparison is presented between the computed settlement in other references and the computed settlement by ELPLA using different Nonlinear analysis methods. The comparison is presented as a cross section at Wing c and tables as in Figure 714 and Table 73, respectively.
The comparison shows that the results of two methods in ELPLA are in good agreement with the calculated results of Russo etc. al. (2013). The second method (LoadSettlement relation as a Hyperbolic Function for LoadSettlement Curve) results are closer to the design results presented by Poulos and Bunce (2008).
Method 
S_{max.} [mm] 
S_{min.} [mm] 
S_{Diff}_{.} [mm] 
Design Values (Poulos and Bunce 2008) 
78 
60 
18 
Russo etc. al. (2013) 
58 
24 
34 
ELPLA – Given LoadSettlement Curve 
58 
29 
29 
ELPLA – Hyperbolic Function for LoadSettlement Curve 
79 
47 
32 
7.8.3 Calculated final pile loads
The maximum and minimum pile loads were obtained from the threedimensional finite element analysis for all loading combinations by Poulos and Bunce (2008). The maximum loads were at the corners of the three “wings” and were of the order of 35 [MN], while the minimum loads were within the center of the group and were of the order of 1213 [MN].
Figure 715 and Figure 716 show pile loads obtained by ELPLA using method: "Hyperbolic Function for LoadSettlement Curve" and method "Given LoadSettlement Curve from pileload test", while Table 74 compares results of max and min pile loads obtained by ELPLA with those of Poulos and Bunce (2008).
Table 74 Comparison between various calculated pile loads
Method 
P_{max.} [MN] 
P_{min.} [MN] 
FEA (Poulos and Bunce 2008) 
35 
1213 
ELPLA – Given LoadSettlement Curve 
38 
11 
ELPLA – Hyperbolic Function for LoadSettlement Curve 
21 
13 
7.9 Evaluation
It can be concluded that results obtained from different analyses available in ELPLA can present rapid and acceptable estimation for settlement and pile loads. This case study shows also that analyses available in ELPLA are practical for analyzing large piled raft problems considering less computational time compared with other complicated models using three dimensional finite element analyses.
7.10 References
[1] El Gendy, M. / Hanisch, J./ Kany, M. (2006): Empirische nichtlineare Berechnung von Kombinierten PfahlPlattengründungen.
Bautechnik 9/06
[2] El Gendy, M. (2007): Formulation of a composed coefficient technique for analyzing large piled raft. Scientific Bulletin, Faculty of Engineering, Ain Shams University, Cairo, Egypt. Vol. 42, No. 1, March 2007, pp. 2956
[3] El Gendy, M./ El Gendy, A. (2018): Analysis of raft and piled raft by Program ELPLA GEOTEC Software Inc., Calgary AB, Canada.
[4] Poulos, H. / Bunce, G. (2008): Foundation Design for the Burj Khalifa, Dubai – the World's Tallest Building.
6th International Conference on Case Histories in Geotechnical Engineering, Arlington, VA, August 1116, 2008.
[5] Russo, G./ Abagnara, V./ Poulos, H. & Small, J. (2013): Reassessment of foundation settlements for the Burj Khalifa, Dubai. Acta Geotechnica (2013) 8:3–15.
[6] Badelow, F./ Poulos, H. (2016): Geotechnical foundation design for some of the world’s tallest buildings.
The 15th Asian Regional Conference on Soil Mechanics and Geotechnical Engineering