Structure/foundation challenges for Australia's Northwest Shelf
Soft rocks, calcareous soils complicate evaluations
Northwest shelf oil and gas developments are characterized by remote fields in a range of water depths and variable soils. Field development studies frequently consider a number of potential sub structure sites to minimize development costs or to complement options for pipeline routing. Detailed costing of these substructure options requires an understanding of likely soil character istics and their impact on the substructure foundation design.
Site investigation, testing, and foundation design techniques developed over recent years have been refined by the work of the Wandoo (1994) and Gorgon Alliances. The Wandoo concrete gravity substructure (CGS) has shown that economic and safe foundation design solutions are achievable in the calcareous soils of the Northwest shelf. Gorgon has refined this approach to develop a significant substructure solution in a manner that will increase operator confidence that economic substructure designs can be delivered.
Whether a gravity based structure or piled foundation solution is to be considered, a well considered framework for site investigation and design should be determined from the outset. This requires an understanding of the geological, geophysical, and geotechnical characteristics of the shelf soils and a testing program that complements the nature of the substructure under consideration.
Calcareous soils are widely distributed on continental shelves in temperate and tropical areas of the world. Similar soils to those of the Northwest shelf are found in the Arabian Gulf and off the coast of India. The soils range from well-cemented calcarenites, through uncemented sands and silty sands, calcareous muds and oozes to high plasticity calcareous "clays." Severe cyclonic conditions prevail along much of the Northwest shelf, setting the region apart from others with similar soils.
Our current understanding of shelf geology has been enhanced by use of carbon dating techniques. The ever-increasing date range over which Carbon 14 dating can be applied augments established methods used to reveal the sequence of geological deposition. When this dating is combined with the knowledge of sea-level changes and geophysical data, a significant improvement in the understanding of material properties can be gained. Advancing and receding ice caps in recent geological time have resulted in sea levels up to 130 meters lower than present. These changing depositional and post depositional (diagenetic) environments have had a significant influence on the engineering properties of these calcareous soils.
During low sea level periods, calcareous rocks form as carbonate-rich water is drawn to the surface by evaporation, depositing the cement calcite. When water levels rise, deposition mechanisms recommence, and the cementation process is retarded and leads to weaker linkages of the intergranular contacts. Under standing both the cement linkages and the nature of the particulate material is essential to predicting behavior on the micro level. On the macro level, recently cemented or indurated carbonate layers may be underlain by, or pass laterally into, uncemented materials, giving rise to variable foundation conditions.
A reliable foundation design requires a detailed site investigation to confirm an understanding of the geology. Additionally, soils data is essential for reliable geotechnical parameters to be established. Exploratory tools and techniques in both the geophysical and geotechnical investigation have to be chosen to be "fit for purpose," so that they complement each other and bring a secure understanding of the ground conditions.
Laboratory testing also forms an important part of the investigation procedure and significant advances have been made since the experiences in the early 1980's. Advances have been made in the difficult problem of in situ density determination and sample preparation at the high voids ratios representative of those found in calcareous soils. Recently, the behavior of cemented calcareous samples is better understood now, particularly as a result of the use of artificial cementing agents to recreate soils near to in situ condition from disturbed samples.
Site investigation planning must be integrated with the development of platform sub-structure concepts at an early stage to maximize efficiency of solution development and ensure that appropriate design parameters can be derived from the site investigation. Advanced soils testing programs seek to carry out laboratory testing at stress levels relevant to those anticipated for the subsequent development.
Foundation design applied to calcareous soils is invariably concerned with brittle behavior. When loaded to low-level strains, the soil response may be relatively elastic. However, as loads increase and cementation begins to break down, pore pressures increase leading to a loss in strength.
With time, these pressures dissipate, but significant volume changes may result before strength is regained. Samples of calcarenite from the Northwest shelf demonstrate these features. At relatively low stress levels, they behave like soft rocks, however, when the cementation breaks down the calcarenite behaves more like uncemented calcareous silt. Stresses at which this breakdown occurs may be in the range of 1-5 MPa. The initially stiff response of a calcarenite sample turns to a brittle failure at around 1000 kPa.
The tendency for these calcareous soils to exhibit brittle behavior or breakdown is relevant to the type of foundation under consideration. Driven piled foundations break down the cemented bonds along the pile shaft during installation, and have high end bearing stresses which generally exceed the soil strengths associated with cementation of the soil.
The piled foundations on Rankin (1980s) and Goodwyn (1992) exhibited related but different facets of this problem. In particular, the degree of loss of skin friction in the calcarenite soft rocks on installation was not predicted and extensive remedial works were required. Research into understanding the reasons for low pile skin friction in calcareous soil is still ongoing and techniques for improving it, for example, through re-calsification of the broken cement bonds are still being developed.
Alternative foundation solutions such as skirted or bucket foundations typically apply bearing pressures onto the soil lower than that required to break down soil cementation. Bucket foundations or suction anchors transfer foundation loads at soil stresses typically in the range 200-1000 kPa, which is significantly lower than piled solutions. Their use will undoubtedly increase where calcarenite layers are not present near the surface.
Gravity structures such as the Wandoo CGS, designed by Arup Energy and installed in 54 meters water depth in 1996, have very much lower maximum design stresses. Peak stresses under the substructure are in the order of 80-100 kPa. At these low stress levels, carbonate materials are not as sensitive to crushing and the foundation solution is readily assured.
Case study: Gorgon
A foundation (FEED) study for the Gorgon LNG project was undertaken during 1998 by Arup Energy as part of the Gorgon Offshore Alliance (Kvaerner, Schlumberger and Clough). Substructure options were developed to support the offshore processing of gas brought from wells located at or beyond the edge of the continental shelf. Dry processed gas was to be exported from the fixed platform to shore for subsequent liquefaction.
An integrated project team was set up to assess various substructure options - concrete gravity, hybrids, jackets with bucket foundations - for potential location in water depths of 80-160 meters. The relative economics of the infield flowline cost, substructure cost, and export pipeline cost would determine the platform location. Multiple costing studies meant that a range of platform sites were considered.
The significant extent of the Gorgon field and planned phased development meant that the substructure could be located anywhere within a 20 km by 20 km area. This represented a unique geotechnical and geophysical challenge, given the sparseness of soil samples recovered from the area.
Extensive detailed geophysical investigations were used to relate known ground stratigraphy to the area of interest. The geophysical, geotechnical and structural engineers worked closely in developing a strategy to identify and then carry out detailed site investigation for the substructure. This included initial selection of prospective platform locations based on economic and foundation considerations.
The geotechnical investigation followed this strategy by firstly carrying out preliminary investigations using wheeldrive piezo cone penetration tests at five sites. Each site was immediately assessed for foundation suitability and ranked to assist in the decision process to select the site for further detailed investigation.
The fifth site at which preliminary investigation was carried out was selected for detailed investigation and the variability across the site assessed using piezo cone tests. High quality push samples were obtained for subsequent laboratory testing.
Complex variation in ground conditions were revealed by the investigations. On representative seismic sections the multiple "B" seismic reflectors are a thick deposit formed during the glacial interval 190,000 to 130,000 years before the present. They represent more heavily cemented horizons, formed when the sea level dropped about 20 meters below the elevation of Site 5, and fluctuated repeatedly, sometimes submerging the site and depositing shallow marine sands, and then rapidly cementing these sands during the a period of exposure.
A deeper channel is part of an old Pleistocene sequence and was cut towards the end of a maximum fall in sea level. It was then filled with carbonate silt prior to the next major rise of sea level. During this rise, between 130,000 to 74,000 years before present, a carbonate sequence formed in relatively deep water. The interval between reflectors "A" and "B" represent this.
The "A" seismic reflectors appear to have formed during the relatively minor glacial stage between about 74,000 and 59,000 years before present, at the end of which sea level fell to just below the elevation of Site 5. Again, cementation with coarsening of the sediment due to the shallowing of the water resulted in Reflector A.
A second channel was cut down into the underlying sequence over a greater width. Filling of the channel commenced with deposition of small cemented fragments of white sands in water depths of less than 10 meters. As sea levels rose, the sediment changed to a very soft mud, and then brown and gray skeletal sand.
A concrete gravity structure in 100 meters water depth was designed for Site 5. It features 14-meter-long steel tipped concrete skirts, which develop most of their required lateral resistance to cyclonic wave loading in the stiff sand layer into which they would penetrate.
The stiff highly cemented sand layer between 12 meters and 17 meters was predicted by the presence of the seismic reflector. In this case, the thickness of the stiff sand layer is sufficient to support the design foundation loads. Dynamic soil/structure interaction analyses were carried out to optimize and confirm the design for this structure. Design maximum bearing stresses on the founding material were limited to only 80 kPa and shear stresses to 20 kPa for the adopted foundation solution.
Gorgon development manager, Dr. Jim Briggs, contributed data from the 1998 investigations to this article.
Fookes, P (1988), "The geology of carbonate soils and rocks and their engineering characterization and description, Engineering for Calcareous Sediments," eds Jewel & Khorshid, Balkema, Rotterdam.
Campbell, K (1999) "Deepwater geohazards: How significant are they," Leading Edge, V18 No. 4, Apr 99.
Carter, J, Airey, D, Fahey M, (1999) "A review of Laboratory Testing of calcareous soils," Research report No. R789, August, Centre for Geotechnical Research, University of Sydney.
Tjelta, T (1998), "Views on Geotechnical Integration," Geotechnical keynote address; SUT Offshore Site Investigations and Foundation Behaviour, London.
Poulos, H, (1999), "Some aspects of pile skin friction in calcareous sediments," Research report No. R786, August, Centre for Geotechnical Research, University of Sydney.
Kucharski, E, Price G, * H, Joer H (1996), "Laboratory Evaluation of CIPS Cemented Calcareous and Silica Sands," Proceedings 7th Australia/NZ Conference on Geomechanics, Adelaide, South Australia.
Humpheson, C (1988), "Foundation design of Wandoo B Concrete Gravity Structure," Offshore Site Investigation and Foundation Behaviour Conference '98, SUT, London.