Deep Foundations on Bored and Auger Piles Van Impe & Van Impe (eds) 2009 Taylor & Francis Group, London, ISBN 978-0-415-47556-3 Design and construction aspects of piled foundations for Eureka Tower Project Jim Slatter & Slav Tchepak Vibropile (Aust) Pty Ltd, Melbourne, Australia ABSTRACT: The Eureka Tower project involved the construction of a 300 m high 92-storey tower, the world s tallest apartment tower at the time, located in Melbourne s Southbank area. An unusual feature of the building is its slenderness, having a height to base ratio of 6 to 1. The construction of the foundations for the project proved to be a challenging task. The geological conditions at the site were complex, highly variable and posed significant construction and technical difficulties, with two layers of high to very high strength basalt above high strength Silurian Siltstone bedrock at a depth of approximately 35 m. The ground water table occurred at 2 m depth and the upper and lower basalt layers were not continuous across the site. To add to the complexity, the loadings imposed on the foundations by the structure were high. The lower basalt provided a suitable founding medium, provided that sufficient thickness was available to ensure that settlements of underlying soils were within acceptable limits. This was difficult to define because of the discontinuous nature of the lower basalt and the variable thickness of that stratum. The foundation solution that ultimately proved to be the most cost-effective was a combination of CFA piles founded on the very high strength lower basalt flow and Bored piles constructed under bentonite drilling fluid founded in the high strength Siltstone when there was insufficient thickness of the lower basalt. This paper discusses the design and construction aspects of the piled foundations for these challenging conditions, including the additional site investigation required to define areas appropriate for each pile type; construction of the piles and the special techniques required to ensure clean bases for the heavily loaded piles; and the testing regime that comprised Statnamic and Dynamic pile loading tests. 1 INTRODUCTION The site for the 92 storey apartment building was originally an industrial area. Because of its proximity to the CBD of Melbourne and the Yarra River, the area rapidly evolved to commercial and apartment usage. Extensive geotechnical work had been carried out for previous developments on the site, which were abandoned, partly because of the costs of providing economical foundation solutions for multistorey developments in the challenging soil conditions. Local builder Grocon proposed a 92 storey apartment building which presented even greater challenges for the structural, geotechnical and piling engineers to provide an economical foundation system given the high loadings that result from such tall structures. The structure had a relatively small footprint, resulting in a height to base ratio of 6 to 1. The top of the tower can flex up to 600 mm in high winds with resulting oscillations being dampened by two 300 ML water tanks on levels 90 and 91. The builder, Grocon, in conjunction with consulting structural engineers Connell Mott MacDonald and geotechnical consultants Golder Associates, issued documentation for the tower foundations comprising bored piles socketed into the strong Silurian bedrock. Initially the piling work was priced on the conforming solutions utilising bored methods, the equipment and expertise for which needed to be imported due to the requirement to penetrate up to 8 m of massive, very high strength Basalt at up to 1.5 m diameter. The costs of the conforming solutions considerably exceeded budget expectations both in terms of cost and program and the project was at significant risk of not proceeding. 2 SITE GEOLOGY The geological conditions at the site are complex and difficult. It is not possible to provide a simple tabulation of soil types with depth, nor will reproduction 323
Figure 1. Section DD. Figure 2. Section FF. 324
Figure 3. Final layout of CFA & bored piles. of bore logs provide a realistic depiction of the geotechnical conditions. The geological complexity of the site was primarily due to two layers of basalt occurring above the Silurian Siltstone bedrock that occurs at a depth of about 35 to 37 m depth. The upper basalt is sandwiched between layers of clay that varied in consistency from soft to very stiff. The upper basalt varied in thickness from zero to about 8 m and was typically of high strength, 80 to 100 MPa and most commonly highly fractured, with fracture spacings typically not greater than 300 mm The lower layer of basalt was generally overlain by dense sands and gravels, occasionally underlain by the same sand and gravel strata or occasionally sitting directly on the siltstone bedrock. The lower basalt was of very high strength, indicated by testing to be in excess of 200 MPa and tended to be massive. The ground water table occurred at 2 m depth. Neither the upper or lower basalts were continuous across the site. The siltstone bedrock was typically a strong rock (up to 80 MPa), with the upper 1 to 2 m being relatively highly jointed before becoming massive in formation. Cross sections from the relatively small site are shown in Figs. 1 and 2 that highlight the variability in soil conditions. 3 DESIGN CONSIDERATIONS Founding piles or pile groups on the upper, highly fractured basalt was not an option because of the high loads, which would result in excessive settlement of foundations due to the compressible soils underlying the upper basalt. By contrast the massive lower basalt provided a suitable founding medium, provided that sufficient thickness was available to ensure that settlements of underlying soils were within acceptable limits. This was difficult to define because of the discontinuous nature of the lower basalt and the variable thickness of that stratum. As a consequence, an extensive geotechnical investigation was performed by the geotechnical engineer to better characterise the condition and extent of the lower basalt with a view to revising the foundation solution. The geotechnical design responsibility remained with the geotechnical engineer, who determined that a minimum of 5 m of strong lower basalt would be required to ensure satisfactory performance of piles founded in that stratum. Where 5 m of basalt was not available, piles would need to be bored through the basalt and socketed into the underlying strong siltstone. A variety of solutions were considered including: An initial proposal to utilise 76 No. 1500 m dia piles socketed into siltstone 110 No. 1200 mm dia bored piles socketed into siltstone Foundation solutions included construction utilising conventional and reverse circulation drilling methods. Interestingly, because of the difficulties of penetrating the very high strength basalts in particular, costs were found to be cheaper for a larger number of 1200 mm dia piles compared to 1500 mm solutions, especially when relatively high strength 325
concrete (70 MPa) was used and piles loaded up to the safe structural limits of the pile shafts for both piletypes. A solution was then proposed by the piling contractor to use high capacity CFA piles founded on the lower basalts where possible, with 1200 mm dia bored piles socketed into siltstone where the lower basalts were less than 5 m thick. This solution was found to be significantly cheaper than all-bored pile solutions and offered considerable savings in construction programme. A final solution incorporating the following was adopted: 243 No. CFA piles 750 mm dia founded on the lower basalt, to a unit pile design (i.e. ultimate) load of 9250 kn (corresponding to a working load of approximately 6800 kn). 28 No.1200 mm dia bored piles socketed in siltstone for ultimate loads of up to 32 MN (working load approximately 25 MN). An allowable stress was adopted as 20 MPa in the basalt. The design of bored piles in the siltstone was done using the program ROCKET to ensure that estimated settlements of those piles would be compatible with bored piles founded in the basalt. Consequently the settlement at the top of the siltstone socket was restricted to 6 mm and resulted in sockets of 4.5 m length. 4 CONSTRUCTION ISSUES A number of issues had to be addressed, including: Proving the performance of CFA piles would be in accordance within specified criteria by load testing those piles. Confirming the constructability of the bored piles, to enable confirmation of production rates and pricing structure of those piles. Confirming that construction procedures for bored piles founded in either the basalt or the siltstone will be satisfactory and that those procedures can be implemented on a routine basis should the CFA trials not meet expectations. In addition to the above technical considerations, the site was also extremely congested due to the multiple concurrent activities which were had to be scheduled in order to meet the extremely challenging construction program. As a result, site management was critically important and the movement of each item of plant needed to be carefully co-ordinated in order to prevent clashes (see Figure 4). 4.1 CFA piles During the course of production piling approximately 75% of the piles required predrilling through the upper basalt layer using crane mounted drilling rigs. After predrilling, the excavation was backfilled with Figure 4. Site congestion was a major concern. cement stabilised sand, ready for construction of CFA piles. Predrilling of clustered piles was be sequenced such that following casting of any pile in the cluster, subsequent predrilling for the next pile would not take place for at least three days. CFA piles were drilled to effective refusal to found on top of the lower basalt and constructed using concrete injection techniques. Due to the sloping surface and negligible weathering profile of the lower basalt, specially designed rock drilling heads had to be adopted to ensure that CFA piles were adequately seated into the very high strength lower basalt. Every aspect of pile construction was fully monitored by on-board computers to ensure the highest quality construction. 4.2 Bored Piles Bored piles required drilling through up to two layers of high to very high strength basalt to socket into the high strength siltstone. Construction under bentonite drilling fluid was adopted as the most economical approach to the difficult conditions. The sockets were formed using conventional rotary drilling methods however extremely high 326
powered machines were required. A 55 tm crane mount drill was required to core through the very high strength basalt (Fig. 5). To facilitate base cleanliness, a series of purpose built tools were developed. Firstly a pilot hole/sump was formed centrally at the base of the 1200 mm diameter socket, a second tool was then deployed to mill a flat surface (or ledge) at the base of the pile (Fig. 6) and finally a third tool was used to sweep any debris on the milled ledge into the sump thus providing a high degree of cleanliness on the load bearing ledge at the pile toe (Fig. 7). The sump, which was formed by coring and chiselling was 600 mm diameter and 600 mm length. The purpose of the sump was to attract base debris during construction, ensuring the remaining 75% of base area to fully utilise end-bearing. Socket walls were grooved in accordance with the requirements of the ROCKET analyses, namely, a minimum 5.5 mm deep by 6 mm wide groove at 100 mm spacing. Piles were concreted within 24 hours of completion of socket drilling, with socket grooving and desanding (to ensure a maximum sand content of 1%) operations carried out on the same day as concreting. Socket inspections by underwater video camera proved the efficacy of construction methods in providing a clean pile base that would satisfactorily support the high loadings. 5 PILE TESTING An initial trial piling programme was instigated to confirm the veracity of the proposed CFA construction technique and to confirm the estimated construction programme. Two non-production CFA piles were constructed and load tested to destruction. While it was understood that all production CFA piles would be drilled to effective refusal, one Figure 6. Ledge, milling & sump coring tool. Figure 5. Coring of basalt. Figure 7. Ledge sweeping tool. 327
test pile was purposely terminated as soon as the auger reached the lower basalt, while the other was drilled to refusal. The purpose of the former was to simulate the potential effects of partial contact with the basalt. 5.1 Bored piles Load tests were not done on any bored piles, however a remote socket inspection device (SID) was used on two bored piles to verify the cleanliness of the milled and swept ledge, and the pile sump. The results of the SID inspections proved that unique construction procedure adopted for the pile toes had met the design objectives and that the ledge had been cleaned to satisfactorily high levels. On this basis the construction procedure for all future bore piles on the project was approved and strict QA procedures were implemented to ensure that these procedures were adhered to. 5.2 CFA piles The two trial CFA piles were subjected to load testing by Statnamic and dynamic methods. Dynamic tests were carried out using a drop hammer of 20 tonne mass after completion the Statnamic tests. The costs of doing Statnamic testing precluded testing of a large number of piles on the project, so the purpose of the subsequent dynamic tests on the trial piles was to gain a correlation with Statnamic and provide confidence in dynamic methods for future routine testing of production piles. The results of the loading tests on the trial piles are shown in Fig. 8. The results of the comparative tests indicated a slightly less stiff response from the dynamic tests compared to the Statnamic tests. However the performance of the piles for both tests was satisfactory. The magnitude of the mobilised loads during the tests was not sufficient to conclusively prove the difference in drilling methods. Although the piles were to be tested to destruction, it was not possible to impart sufficient energy during the tests to damage the piles. Pile Top Load (kn) 25000 20000 15000 10000 5000 0 Figure 8. Pile No 6 0 10 20 30 40 50 Displacement (mm) DLT 128kJ Statnamic UPM Pile Top Load (kn) 25000 20000 15000 10000 5000 0 Pile No 5 0 10 20 30 40 50 Displacement (mm) DLT362kJ Statnamic UPM Comparative statnamic and dynamic load tests. During routine piling an additional 5% of the piles were tested dynamically, with all results between the bounds of the DLT results indicated in Fig. 8. All parties were satisfied that the load testing regime was satisfactory given the level of QA available from the computerised monitoring of every CFA pile. 6 CONCLUDING REMARKS A cost-effective and innovative solution for the foundations for the Eureka Tower project was brought about by co-operation between consulting structural and geotechnical engineers and piling contractors in what were extremely difficult geotechnical conditions. The application of different pile types and construction techniques was novel and innovative and resulted in the final cost of the foundations being reduced by over 30% and the program reduced by approximately 3 months. The construction was performed in difficult circumstances compounded by the restricted space available on site. In this manner, the skills of all parties were utilised, to the benefit of the project. REFERENCES Seidel, J.P. 2000. Rocket 3.0 Manual. Monash University. Melbourne. Golder Associates. 2001. Report on geotechnical investigation. Eureka Tower. South Bank. No 01615054. 328