The successful installation of the Europipe 16/llE Riser jacket in July 1994 and the coming installation of Sleipner T jachzt in spring 1996 proves that suction installed skirted plate foundations (buckets), may not only be a competitive alternative to piles, but also a complementaryfoundation solution both for noncohesive and hesive soils. This paper presents the background for the d.exign analyses of the Europipe 16/llE and Sleipner T foundations in terms of the performed field and model tests, as welt as the theoretical models and practical analysis tools utilized for design. INTRODUCTION This paper is a joint presentation made by the designers of the Europipe 16/1lE jacket, Aker Engineering, and thedesigners of the Sleipner T jacket, Kvaerner Ed and wright. The active development of bucket foundations for steel jackets was initiated by Statoil's chief foundation engineer Tor Inge Tjelta and lead structural engineer Morten Baxheirn at the end of 1989. The successfid installations of the Veslefrikk jacket, the deep skirted concrete platform Gullfaks C and the suction installed tether References included at end of Part 1 and Part 2 anchors at Snorre played a vital part in this development. However, soil conditions for the two jackets in question are dominated by dense to very dense sands, while the previous skirted foundations were installed in soft to from clays. Two novel and important issues were identified as critical with respect to the design of the bucket foundation namely.Penetration of the skirts through the very dense sandTension capacity, particularly under cyclic loading To assist in the development of practical design tools, Statoil initiated field testing comp-rising penetration and capacity tests at the actual offshore sites in late 1992. These tests formed the basis for the Europipe platform bucket design which was completed in late 1993. At the end of 1993, a comprehensive model test program was initiated aimed specifically at improving the knowledge and understanding of tension load capacity and clic loading effects for inclusion in the Sleipner T bucket design. Also included in this program were a number of penetration tests for verification of the penetration analyses performed for the Europipe buckets. These tests have been filly utilized during detailed design of the Sleipner T platform. The design is now complete, and fabrication of the jacket is under way. In the following, the background and design of the Europipe 16/1lE and Sleipner T bucket foundations are presented separately. Key elements from field and model testing arc described and mechanisms of behaviour are presented. This experience is coupled with theoretical models and transformed into design tools for the prototype foundations designed for the Europipe 16/1lE and Sleipner T platforms. PART 1- EUROPIPE 16/llE Key design data Europipe 16/1lE is a riser platform for the Europipe transport system, situated in Block 16/11 in the Norwegian sector of the North Sea. Key parameters for the site and structure are water depth 71 m, topside weight 5000 and dry support structure weight 3800 t. A 3D view of the structure is presented in Fig. 1.1.

Anchor caissons, or suction anchors as they are often referred to because of the method of installation, are an attractive alternative to conventional high-capacity drag anchors in soft seabed sediments. The paper is concerned with the use of such anchors to withstand quasi-horizontal loads, such as are applied through catenary mooring chains. A new threedimensional upper bound analysis is described, for calculating the limiting geotechnical capacity of the anchor for different depths and angles of chain attachment. Results are also presented from centrifuge modelling of an anchor of 5.4 m diameter, and embedment ratio of 2.3, installed in fine-grained calcareous soils with a high proportion of clay-size particles. The paper addresses issues such as the extent to which suction occurs on the trailing edge of the anchor, the rapid reduction in capacity (and change in motion of the anchor) that occurs if that suction is lost, the vertical and horizontal capacities, and the response under monotonic and cyclic loading with a typical chain configuration. Introduction Suction caissons are being used more widely as alternatives to drag anchors, offering various advantages (Colliat et al, 1997):the vertical and horizontal capacity of suction caissons can be defined more precisely than for drag anchors, in terms of conventional soil strength parameters;suction caissons are relatively straightforward to install, with no requirement to proof load to a high proportion of their design loads;the final location of suction caissons on the seabed is fixed, allowing well-focused site investigation, whereas for drag anchors uncertainty in the required drag-length leads to uncertainty in the final anchor locations. To date, the geometry of suction caissons has tended to be relatively low aspect ratio, with lengths less than 3 times the diameter. However, higher aspect ratio caissons can be installed by combined deadweight and suction, and may offer improved anchoring capacity for a given weight of caisson, particularly where the soil strength increases with depth. The main focus of this paper is on the capacity of suction caissons when subjected to quasi-horizontal loading from a mooring chain. This type of loading is rather different from the quasi-vertical loading of suction caissons for TLP anchorages (Dyvik et al, 1993; Anderson et al, 1993; Clukey et al, 1995), and has received rather less attention although small field trials have been reported by Keaveny et al(1994). A three-dimensional upper bound analysis is presented, based on the flow mechanism proposed by Murff and Hamilton (1993) for laterally loaded piles. This mechanism is for a single cylindrical caisson, although it can also be extended to pairs of caissons or square groups of four caissons. The attachment point (or pad-eye) for the anchoring chain is generally attached at a depth of 50 to 70 % of the total embedment of the suction caisson, in order to optimise the lateral resistance. This means that the anchoring chain will subtend a small angle to the horizontal (typically 10 to 20°), as shown in Figure 1, imparting a significant vertical component of force.

This paper describes the important geotechnical principles relating to the installation of bucket foundations in dense sand, where 'suction' pressures are applied to both increase the static driving force, and more importantly, to degrade the penetration resistance. The paper will present a series of finite element analyses that demonstrate various important mechanisms. These will be compared with results from an extensive series of model tests and with data collected during the installation of the bucket foundations for the Draupner E platform (formerly Europipe 16/11E). Introduction In order to penetrate the skirts of a bucket foundation into the seabed, a 'suction' operation is performed. During this process, pumps are used to evacuate water from within the sealed skirt compartments, which creates a differential water pressure. This has two beneficial effects in sand soils:Water seeps down and around the skirt tip, and then upwards within the skirt compartments towards the base plate. Given sufficient time, approximately steady state seepage gradients will form (complete steady state conditions never develop since the skirt continuously penetrates). The downwards seepage gradient outside the skirt, acts to increase the effective stresses in the soil and consequently to increase the external skirt friction. Conversely, the upward seepage gradient inside the skirt acts to reduce the effective stresses in the soil. This reduces the internal skirt friction, but most significantly, degrades the skirt tip resistance. The net effect of these processes is a substantial reduction of the total penetration resistance.The differential pressure created across the base plate by the 'suction' operation provides an additional force that helps drive the skirts to the desired penetration. In addition to these two beneficial effects, one important and potentially detrimental effect must also be considered. The degradation process described occurs because the effective stresses within the skirt compartment are reduced by the upward seepage gradient. However, this gradient is limited since the effective stresses can never be less than zero. The onset of this state occurs at a 'critical gradient', and is quite commonly referred to as a 'quick' condition. It is defined by:(Mathematical equation) (Available in full paper) In related applications, soil liquefaction or 'boiling' has been observed at this state, and this is quite often accompanied by the formation of 'piping' channels. For bucket foundations, these phenomena appear highly undesirable. Extensive soil liquefaction would cause a large amount of soil heave which could impair skirt penetration but, more importantly, would seriously affect the in-place foundation performance. The formation of 'piping' channels would cause a breakdown in the hydraulic seal across the base plate that would halt, at least temporarily, further penetration. Given these consequences, it could be considered that a large safety factor against the formation of 'quick' conditions should be specified. However, from the model test results described in subsequent sections, it will be shown that very large internal gradients were nearly always required to cause skirt penetration and in some cases, the critical gradient was reached. However, except where deliberately provoked, extensive soil heave did not occur, nor were piping channels formed. This suggests that some in-built safety mechanism acts to prevent these unstable phenomena.

The Laminaria FPSO installed in 1999 represented the first use of suction piles in the calcareous soils offshore Australia. This paper will describe aspects of the site investigation performed, it will discuss the a-prioi predictions for the suction pile installation characteristics and will present a detailed back analysis of the actual measured data obtained during the suction pile installation. In light of these findings, centrifuge model tests reported previously at OTC will also be reinterpreted. The implications of the installation characteristics on the in-place performance will also be summarised. Introduction The Laminaria/ Corallina field is located in the Timor Sea in the Zone of Cooperation between East Timor and Australia. In October 1998, nine suction anchor piles were installed at this location to provide the mooring for the Northern Endeavour FPSO (installed in 1999) on behalf of Woodside Energy Ltd (WEL) and its partners. This is believed to represent the first use of suction anchor piles in a predominately carbonate material. The anchor piles were designed by a Kvaerner - Single Buoy Moorings consortium (KSC) and installed by Coflexip Stena JP Kenny (CSK). Advanced Geomechanics (AG) was retained by WEL to organise the site investigation and to provide the geotechnical interpretive report, to verify the design of the suction piles and to provide an interpretation of the measured installation data. Historically, carbonate soils have proven to be problematic in offshore foundation engineering2. However, most of the previous problems have been concerned with uncemented carbonate sands and calcarenites, where very low skin friction may be mobilised for driven piles. At the Laminaria location the soil comprised carbonate mud, which superficially appeared very similar to normal highly plastic clays. For a predominately laterally loaded anchor pile, there was little concern as to the expected lateral resistance behaviour, and this was further documented through centrifuge model tests1. However, even for this type of anchor, an axial load component is applied due to the inclination of the chain at the buried padeye and also due to installation tolerances. The Laminaria soil appeared to exhibit several "clay like" properties and consequently there was some expectation that traditional pile design rules such as those in API3 could have been applicable in this case. However, being acutely aware of the extensive problems experienced with Australian carbonate soils in the past and with conflicting data available, it was believed that a less optimistic approach was appropriate. At one stage KSC proposed using a temporary lid for the suction piles but, after discussion, agreed to adopt fixed permanent lids. It transpired that this conservative approach was well rewarded. This paper presents the story of the site investigation (which also included the first commercial application of the T-bar), presents key aspects of the suction anchor pile design and presents the installation data and an interpretation of the findings. A re-evaluation of the original centrifuge tests1 is also provided.

This paper presents an original analysis of the flotation potential for a pipeline buried in soil which may not be fully liquefied, but nevertheless exhibits a significant excess pore pressure induced by wave-induced soil shearing.For pipelines that are light enough to be subjected to flotation, the accumulation of pore pressure enhances the flotation potential in two ways:• The pore pressure accumulation in the seabed increases the upwards buoyancy force exerted on the pipeline.• The increase of excess pore pressure in the soil reduces the soil resisting force which restrains the pipeline to prevent flotation. Ultimately, the excess pore pressure may become equal to the initial effective stress level, which implies full liquefaction. It is commonly accepted that this implies a high risk of pipeline flotation if the submerged specific gravity of the pipe is less than the total unit weight of the soil. However, the combination of an increased buoyancy force and a reduced soil resistance as excess pore pressures accumulate mean that the buried pipeline may reach an unstable state before full liquefaction is reached. Hence pipelines may also float in soils that are not fully liquefied. This paper will present the theoretical background for assessing both the uplifting force exerted on the pipeline and the corresponding soil resisting force, as a function of both the excess pore pressure level in the seabed and the cover depth over the crown of the pipeline.The model used for assessing pore pressure accumulation in the seabed is also described and then an example application is presented, comprising a typical pipeline in a backfilled trench subject to a design storm.