Dr James D Murff, or ‘Don’ as he prefers to be called, was born in 1941 in Houston, Texas, and raised in West University Place when it was a small bedroom community on the outskirts of town. In 1963 he obtained a Bachelor of Science and Engineering from the US Military Academy at West Point. He joined the US Army Corps of Engineers and served in Texas, Panama and Vietnam. His meritorious service in South Vietnam earned him the Bronze Star Medal and the Army Commendation Medal. Upon his return to the USA, he attended Texas A&M University from which he obtained MS and PhD degrees in 1970 and 1972. He then joined the offshore division of the Exxon Production Research Company in Houston, Texas. Throughout the years, he advanced through various levels of responsibility in offshore geotechnical and earthquake engineering. In this capacity he was responsible for planning, conducting and implementing geotechnical research; developing geotechnical specifications and design methodology for major Exxon platform installations; and consulting and troubleshooting for Exxon affiliates. He eventually achieved the position of Senior Research Advisor, one of Exxon's highest technical levels. Until his retirement in 1999, Don was the goto expert and ultimate authority for all geotechnical matters for the largest publicly traded oil and gas company in the world. If Don said, ‘No’, they would not do it; if Don said, ‘Yes’, everybody felt good about doing it. In parallel with his career at Exxon, Don was very active in industry committees, particularly the American Petroleum Institute (API) Geotechnical Resource Group, of which he was a member from 1976 to 2001, and the Chairman from 1978 to 1985. He led the research project on centrifuge research for offshore piling with Prof RF Scott.

ABSTRACT This paper describes a range of techniques for estimating the capacity of offshore foundations, which has always been a main issue in foundation design and remains so today. The paper, however, takes a tack that is slightly out of the mainstream - by emphasising methods of plastic limit analysis rather than more traditional approaches. It begins with a brief history of offshore geotechnical developments, describing how design methods have evolved for shallow foundations and pile foundations, and the types of loads, site conditions and foundation geometries encountered. A number of simple solutions are provided with detailed example problems. This paper proposes that plastic limit analysis methods have the potential to supplement and enhance more traditional methods. 1. Introduction I am sincerely honoured to be invited to give the inaugural McClelland Lecture. I am humbled by the task before me as I sincerely wish to produce something that Bram McClelland would have appreciated. At first I leaned toward a subject that more characterised his expertise and interest - engineering geology, site investigation and foundation design. On further reflection however I concluded that Bram delighted in developing engineers who followed their own interests, not in his image, but in their own unique ways. That is the kind of leader he was. This epiphany led me to select a topic that has long been a passion of mine - bridging the gap, sometimes chasm, between theory and practice. I believe this is what he would have wanted from me. Estimating foundation capacity has always been a central issue in foundation analysis and design. Various different methods are employed in this practice, many of which involve ad hoc assumptions and empirical models. This paper focuses on one such advancement, plastic limit analysis (PLA), a methodology that is theoretically sound, internally consistent and surprisingly simple to apply.

ABSTRACT This paper reviews some key issues regarding the cyclic loading response of offshore piled foundations. Starting with axial loading it considers: the cyclic loading that can be expected; the fundamental responses of piles driven in clays and sands; frameworks for understanding axial cyclic response and specifying cyclic soil testing; and approaches for practical application in design. The review then moves to consider pile responses to moment and lateral loading, distinguishing between flexible and relatively rigid piles and anchors. A range of possible design approaches is considered and it is argued that current routine practice needs to be reconsidered. Practical methods now exist to address the potentially highly significant effects on axial capacity of piles that experience high ratios of cyclic to average loads. New research and calculation procedures are emerging that offer significant improvements in a broad spread of topics. 1. Introduction Interest in the behaviour of piles under cyclic loading grew in the 1980s to meet challenges posed by inherently fail-unsafe Tension Leg Platforms (with the first TLP being installed at Hutton in 1984) and heavily loaded deeper water fixed platforms, such as the Cognac jacket set in 320m water. Briaud and Felio (1986) assembled for API a database intended to resemble fine marine sediments covering the cyclic behaviour of clays in: laboratory tests, cyclic model experiments and axially cyclic field pile tests. They considered 16 studies on piles with diameters greater than 150mm, most of which were strain-gauged to measure axial load distributions. Local shaft friction, pore pressure and radial stress measurements were attempted in some cases, although these parameters are notoriously hard to sense reliably. The response of piles driven in sands was not addressed. The piles were submitted to significant numbers of load cycles (typically 100 to 1,000) with frequencies generally around 0.1 Hz.

ABSTRACT Rising gas occurs naturally and as a result of drilling, injection and construction activities, such as air jetting, fracturing, jet grouting and tunnelling. The effect of rising gas is difficult to quantify or model, but is known to reduce the bearing capacity of both shallow and deep foundations. Under gravity based structures (GBS) it can cause problems of flotation and subsequent base sliding and internal over pressure. Storm waves can liquefy shallow gaseous deposits. Gas venting facilities are essential but not always provided on offshore structures. The paper discusses rising gas in terms of sources, pathways, flow calculations, gas experiments trials, and calculations, settlement effects and precautions, as well as its effect on GBS, on bearing capacity (especially during storms) and on shallow and deep foundations. 1. Sources of Gases The most obvious sources of gases are hydrocarbons rising naturally or as a result of drilling operations. Additionally gases, such as nitrogen, are injected as liquid into strata to increase well production, while other gases are injected for storage or disposal. Compressed air is used in tunnels, caissons, jet grouting and jetting. There are also volcanic gases, hydrates and organic deposits releasing methane. 2. Flow Pathways Pathways for rising gas are mainly dependent on the lamination and dip of the strata and local anomalies such as faulting, joints or weaknesses. Layers of gas can be trapped below individual laminations or cyclotherms (due to seasonal depositional changes in particle size). Seismic surveys show numerous strong ‘ghost’ reflections, which correspond to the pattern of lamination but cannot usually be distinguished in recovered core. A likely explanation is that trapped gas provides the ghost reflection. The air lifts the cuttings, but can also escape laterally and appear in adjacent boreholes. Shot hole blasts can show a similar behaviour.

ABSTRACT SOLCYP is a research and development project conducted in France to: understand the physical phenomena conditioning the response of piles to vertical and horizontal cyclic loads; develop advanced design methods; and initiate pre-normative development of methodologies that may later be included in national and international codes or professional standards. The potential applications include conventional structures, such as electricity pylons or chimneys, high rise towers and high speed train bridges. However, a central emphasis is also given to more novel foundations for offshore and onshore renewable energy engineering. The paper describes the objectives and overall technical content of the project. Several companion papers focus on more specific aspects and the results obtained so far. 1. Introduction The oil and gas industry has developed various procedures for considering the effects of large wave cyclic loads on foundations for offshore structures. Design guidelines include the American Petroleum Institute (API) RP 2GEO (2011); Det Norske Veritas (DNV) Foundations (1992) and the International Organization for Standardization (ISO) 19901–4 (2003). In addition, the offshore turbines industry is progressively adapting such methodologies given in DNV-OS-J101 (2011) and Federal Maritime and Hydrographic Agency (BSH) publication entitled, Design of Offshore Wind Turbines (2007). Surprisingly, the effects of cyclic loading on foundations are largely ignored in most civil engineering and building activities. French codes and Eurocode7 (2007) reflect this poor level of consideration. A committee working under the umbrella of the French national agency, IREX, called for national engagement in an ambitious research and development project to address the present lack of guidance regarding piles under cyclic loading. This paper describes the objectives and overall technical content of the project. Several companion papers focus on specific aspects and initial outcomes. Cyclic loads may be essentially environmental (e.g. wave, wind) or operational in origin.

Abstract Offshore shallow foundations may be subjected to uplift due to overturning or buoyancy loading. Peripheral (and often internal) skirts can enable transient tension loads to be resisted because of negative excess pore pressures developed between the underside of the foundation top cap and the soil plug confined by the skirts. Uncertainty exists with regard to the duration over which these negative excess pore pressures can be maintained and the effect of a gap forming along the skirt-soil interface on the transient and sustained holding capacity. This paper presents results from drum centrifuge tests carried out on a shallow skirted foundation subject to transient and sustained uplift in a lightly overconsolidated clay. Results from baseline tests with an intact skirt-soil interface are compared with tests in which a gap was created along the skirt-soil interface prior to transient and sustained uplift. The results are promising, showing for example that uplift loads of 40% of the peak undrained capacity were maintained for up to two years without significant foundation displacement when an intact foundation-soil interface was maintained. However, they also reveal that the presence of a gap may halve the time to reach similar displacements. Introduction Shallow skirted foundations are widely used offshore to support small platforms, seabed protection structures, storage tanks, and subsea frames for oil wells and pipelines, as well as for larger fixedbottom and floating structures (e.g. Støve et al., 1992; Tjelta, 1994; Bye et al., 1995; Watson and Humpheson, 2007). Skirted foundations are also an attractive option for mooring or supporting current meters and wind turbines offshore. Skirted foundations may comprise a top plate a peripheral skirt, and sometimes internal skirts or a cluster of individual skirted units connected together. Regardless of configuration, the foundation penetrates the seabed, confining a soil plug inside the structural members.

Abstract The performance of a hybrid foundation for subsea systems, defined as a shallow skirted mat foundation featuring short piles at the mat corners, has been investigated through a series of centrifuge tests. The foundation, with and without corner piles, was subjected to eccentric monotonic loading along the x , y and z axes, resulting in combined loading over 6 degrees of freedom. Modes of yielding were identified and the contribution of the corner piles to the bearing, sliding, overturning and torsional capacities of the hybrid subsea foundation was quantified. Results revealed that (a) centrifuge modelling could capture the strain hardening arising from the plunging nature of the mat foundation yield; and (b) the addition of the corner piles to the shallow mat resulted in a change of yielding mode from shearing at the mat invert to overturning, with a significant increase in the foundation capacities. Consequently, corner piles appear to be an efficient option to reduce the size of the subsea system foundation required to withstand a given set of combined loading. 1. Introduction Subsea mats are used in deep waters as foundations in soft normally consolidated (or lightly overconsolidated) clay to support facilities such as pipeline terminals, jumpers, riser bases and manifolds. They are typically subjected to various combinations of loading in all six degrees of freedom. In some cases, a mat may not provide sufficient resistance against bearing, sliding or overturning failure. The use of pinned piles in each corner of the mat may then be considered to increase the sliding and overturning capacity of the foundation. Such a foundation is defined here as a hybrid subsea foundation. Hybrid subsea foundations have already been deployed in situ , but neither formal guidance nor experimental data existed at the time to assist in their design.

Abstract The vast majority of offshore wind farms constructed to date are supported on monopile foundations. These monopiles consist of a large diameter (>4m) open-ended steel piles driven into the seabed to a specified penetration. While laterally loaded piles have been used for many years in the offshore oil and gas sector, they typically have diameters below 2m and a slenderness ratio (ratio of pile length to diameter) in excess of 20. In contrast, monopiles used in the offshore wind sector typically have slenderness ratios of 5 to 8. Design methods developed for relatively slender flexible piles are unlikely to provide accurate predictions of the response of more rigid monopiles to loading. This paper presents the results of a field test performed on an instrumented monopile installed at a dense sand research site in Blessington, Ireland. The pile, which had an external diameter of 340mm, was driven into the dense sand to a slenderness ratio of 6. It was also instrumented with 11 levels of strain gauges to capture the load transfer and bending moments along the shaft. The load test results show that conventional design procedures, such as the Det Norske Veritas (DNV) or the American Petroleum Institute (API) approaches, grossly underestimated the lateral capacity of the monopile. At the end of the paper, a 3D finite element analysis of the pile load test is presented. 1. Introduction There has been a significant drive to develop offshore wind energy resources in the last 20 years. To date, most offshore wind turbines (>75%) have been constructed on monopile foundations in relatively shallow water (<30m). Monopiles are large openended steel tubes, with typical diameters ( D ) that are between 4m and 7m, and embedded lengths ( L ) typically less than 30m. This gives a resulting slenderness ratio ( L/D ) of between 5 and 7.

Abstract With the development of deep-sea resource exploration, long steel-pipe piles of large diameter are widely used in offshore platform constructions. Some steel piles may have a diameter that exceeds 2m, and length that exceeds 100m. For such piles, it is almost impossible to accomplish continuous pile driving, as it may be affected by complicated environmental conditions of the sea, the welding of pile connections and the hammer replacement. After a long-term cessation of pile driving, it might become harder to resume and sometimes refusal can even happen. Analyses showed that the soil permeability, the crack development in the soil surrounding the pile, the duration of cessation and the soil-plug effect in the pile pipe are the main factors affecting the soil strength and skin friction. Based on the effective stress principle and the cavity expansion theory, a method has been proposed to evaluate the excess pore water pressure in the soil surrounding the pile during pile driving. In addition, the 1D wave equation has been introduced into the pile-driving analysis. The soil properties in the pile driving analysis are modified in terms of the measured data until the calculated results match the driving records. Therefore, the pile bearing capacity after refusal can be eventually determined. 1. Introduction Long steel-pipe piles of large diameter have been widely used in offshore engineering. Due to the complexity of working conditions offshore, factors such as weather conditions, ship arrangement and hammer replacement, may prevent piles (deep penetration pile) from being driven to the design elevation continuously, and driving interruptions can take place. In this paper, resuming driving after driving cessation is called re-driving. It can take hours, days or even a few weeks to restart the pile-driving process. Long-term cessation can lead to excess pore water dissipation and soil strength setup.

Abstract A new Interstate 10 (I-10) Twin Span Bridge over Lake Pontchartrain was recently constructed to replace the old bridge that was heavily damaged by Hurricane Katrina in 2005. A large portion of the bridge is supported by batter pile group foundations. To evaluate the performance of batter pile foundations under lateral loading, a selected pier (M19 eastbound) of the new bridge was instrumented and used to monitor the pier during a unique full-scale lateral load testing. The M19 pier foundation consists of 24 precast prestressed concrete (PPC) 33.53m (110ft) long batter piles, among which 8 piles were instrumented with microelectromechanical sensor (MEMS) in-place inclinometers (IPI), and 12 piles were instrumented with strain gauges. The test was conducted by pulling the M19 eastbound and westbound piers toward each other by using high-strength steel tendons. A maximum of 8320kN (1870 kips) lateral load was applied in increments. A high-order polynomial curve fitting method was applied to fit the measured rotation profiles from the IPIs. The fitted rotation curves were then used to deduce the bending moment, shear force and soil reaction profiles. The calculated moments from curve fitting were compared with the moments calculated from strain gauges, and the results showed good agreements. The p-y curves of the soils at different depths were back-calculated, and the results showed little evidence of group effect. 1. Introduction Pile foundations are usually designed to support highway bridges and other structures, primarily to safely carry superstructure axial loads deep into the ground. However, in many cases structures (such as bridges, quays and harbours) are subjected to lateral loads caused by high winds, wave action, water pressure, earthquakes and ship impacts. Therefore, it becomes essential to understand the resistance behaviour of piles and pile groups to lateral loads.

Abstract Current offshore foundation technology is being transferred successfully to the renewable energy sector. Still, there is clearly scope for developing foundations that are more tuned to the needs of the renewable power systems such as wind turbines. One such approach is the hybrid monopile-footing system with a proven record of improving the ultimate lateral resistance, particularly in cohesionless soils. This paper builds on to the previous studies by investigating the behaviour of the hybrid system, such as the effect of footing size, the magnitude of pre-loading and its significance in developing sufficient contact pressure beneath the footing, and the importance of the degree of rigidity. 1. Introduction Due to the needs of ongoing developments in the oil and energy sector, the design of offshore foundations is constantly evolving. In the hydrocarbon extraction sector, exploration and development is moving into ever deeper water, resulting in extremely challenging geotechnical conditions. The development of sites for offshore wind farms (such as Round 2 and 3 in the UK) is also extending into deeper water. The increase of wind turbine generator capacity is requiring significant development in foundation design to generate economic and practical solutions to the installation of these deepwater wind farms. Offshore foundations are generally subject to combined loading conditions consisting of self-weight of the structure ( V ), relatively high horizontal loads ( H ) and large bending moments ( M ). The preferred foundation system to date has been the monopile, which has been successfully employed for the majority of the offshore wind turbines installed. The advantage of the monopile is that it can be installed in a variety of different soil conditions even when loading conditions are very high. For instance, in many of the proposed offshore wind farm locations superficial seabed deposits are often underlain by weak rocks, such as mudstones and chalk.

Abstract Accurate spudcan penetration prediction is important for the safe and economic management of mobile offshore jack-ups. However, the current method for dealing with loose sand - by selection of a low friction angle in a general shear failure analysis - is unsatisfactory because the material also experiences significant compression under load. A new simple method is proposed that includes compression and allows spudcan penetration analysis to be based on actual measured site soil parameters. The method is explained, guidance given on site investigation requirements and illustrative case histories documented. The new method's predictions are a good match to field records. 1. Introduction An important aspect of the safe management of jackup rigs is the provision of accurate leg penetration predictions, with a narrow uncertainty range. When predictions consistently match the eventual site outcomes, they become a resource that is consulted and relied upon by the rig move team. This means that discrepancies between prediction and actual leg penetrations will be taken seriously, and feedback sought from the geotechnical engineer on the appropriate response. While much attention is given to making accurate predictions at soft clay and layered soil sites, where leg length and punch-through are issues, less attention is given to sand sites. At a dense sand site this is understandable, as spudcan penetration tends to be small. However, in loose sand the same approach can grossly underestimate the penetration recorded in practice. This also has implications for the prediction at a punch-through site, where the crust is loose rather than dense sand. The most common technique used by geotechnical engineers to deal with loose sand is to select an artificially low friction angle (φ) that, when used in the conventional bearing capacity equations, matches the penetration results at previous sites.

ABSTRACT On a North Sea major gas producer, high pressure has been observed in the outermost annulus of several wells and anomalous pressures were observed below the gravity platform base (inside skirt compartments). Gas was suspected to migrate along well-paths, into well annuli and, in the worst case, past the conductor shoe and continue to the surface and platform foundation. A detailed study was initiated to investigate the gas source, gas migration mechanism and the pressure building component in a system that was believed to be shallow gas driven. The gas migration as part of natural gas flux in the seabed and the relation to formation of pockmarks in the Norwegian trench was investigated as part of this study. The work has provided a good understanding of shallow gas migration below a production platform and has identified the pressure building mechanisms observed in wells and foundation. They are all explained by natural seabed processes and no concern exists for the long term safety of the platform. INTRODUCTION The Troll A Platform is a huge concrete gravity-based structure (GBS) located in the Norwegian trench at a water depth of 305m (see Figure 1 and Hansen et al. 1 ). The platform was installed in 1995 and, in the following two years, 40 wells, 39 gas production wells and one monitoring well were installed. A typical well design is shown in Figure 2. The platform is the North Sea's largest gas producer with an average yearly production in the order of 26 GSm 3 (923 billion ft 3 ). After a few years of production it was realised that pressure bleed-off activity from the outer annuli was high, for some wells 10?20 bleed-offs every month. It is standard practice to set safe threshold values for pressure in well annuli and, if this value is reached, pressures are reduced by bleeding off gas. The threshold for the outer annuli was 6 bar, and this was raised to 15 bar after an evaluation of wellhead seals in 2001. This action reduced the problem of very high bleed-off activity to more normal conditions, typically to once a month for an average well. High annulus pressure is frequently observed in oil and gas production wells both onshore and offshore. The pressures are mostly associated with small amounts of gas seeping along the well-path and entering into the annulus. Cement between the casing and the formation or between two casings doesn't stop this migration process 2 . In parallel to the problem of high well annuli pressures, high pressures in the foundation skirts had been observed (Figure 3). High pressures had been expected here, since the Troll A structure is heavy and sits on soft soils. Also the foundation solution, 36m-long cylindrical concrete skirts that displaced the soil during installation as they were pushed into the seabed, created high pore pressures in the soil and skirt compartments. The combined pressures from platform weight and soil displacement by concrete skirts are a result of foundation consolidation causing increasing pressure at the foundation base

ABSTRACT An attempt of using a T-bar to predict spudcan penetration resistance in clay has previously been proved feasible. This methodology enables the effects associated with the spudcan large penetration on its resistance to be accounted for. To extend the application of miniature penetrometer to assess spudcan penetration in two-layered soil, additional effects such as normalised spudcan level to the soil interface and soil wedge trapped underneath the spudcan base require in-depth investigation. Based on some preliminary results of spudcan penetration tests performed in different layered soil profiles, this paper suggests a means to incorporate these effects into the design of spudcan bearing capacity. Introduction The design of spudcan foundation is usually considered analogous to that of the conventional preembedded shallow foundation despite its large diameter, conical base geometry and large penetration mechanism. In recent years, research investigating the size effect 1 and conical base effect 2,3,4 on the spudcan bearing resistance have been carried out with the aim to enhance the confidence level of using existing theoretical solutions. Nonetheless, the acknowledgement of the continuous change in spudcan foundation failure mechanism with penetration 5, 6 unfolds the inherent limitations underlying these solutions, which were established based on a single mode of failure mechanism. Since they share similar penetration mechanism, this leads to the idea of utilising the readings of miniature penetrometers ? such as T-bar, ball and cone ? to shed lights on the spudcan penetration behaviour. To adopt this alternative approach, the understanding of the performance of spudcan and miniature penetrometer in different types of soil profiles and how they differ from each other is to be established. Background The concept of using miniature penetrometer to predict spudcan bearing capacity combines the knowledge developed in site investigation and foundation design. This concept has been implemented in pile design and the understanding of the concept has been considerably well developed 7, 8 . Owing to the very large dimension and uncommon base geometry of the spudcan, compared to those of site investigation penetrometers, more research is needed to apply this concept to spudcan design. In clay, the use of T-bar is becoming commonly used due to its simplicity in data interpretation and advancement in quantifying the clay properties. Owing to the differences in penetration rate, dimension and geometry, the variations that could possibly exist between T-bar and spudcan penetration resistances are mainly due to the effects of drainage and strain rate on clay characteristic, as well as soil wedge trapped underneath the spudcan 9 . On the other hand, piezocone and ball are deemed more suitable to penetrate through sand due to structural compatibility. Considering the huge dimension of spudcan and its limited penetration in sand, the stress level induced strength variation can vary the piezocone and spudcan resistance profiles developed in the sand to a considerably large extent. To extend the penetrometer based design approach to spudcan foundation in layered soil, additional aspects are to be considered, which create more challenges to the whole process. Based on limited studies performed using miniature cone penetrometer 10, 11 and spudcan 12, 13

ABSTRACT This paper reviews the research and debates that have led to substantial changes being made in 2007 to the API-RP2A recommendations for assessing offshore driven pile axial capacity. The reasons for the conventional Main Text method's large scatter, strong skewing and significant biases are explored, with particular emphasis on piles driven in sand. Recent alternative design frameworks are reviewed critically and conclusions are drawn regarding their practical application. Comments are also made on predicting load-displacement behaviour, assessing the impact of load cycling, group interaction effects and aspects of foundation disturbance by drilling. INTRODUCTION The technologies associated with the manufacture and installation of offshore piles are relatively mature; very large piles may now be driven routinely in a wide range of water depths and geotechnical settings. However, the understanding of the ground's reaction to driven pile installation and loading has lagged behind the impressive developments made by the offshore construction industry, as design approaches are still in an imperfect state of evolution. Severe problems have arisen during pile installation in some major projects1. Considerable mismatches have been found in other cases where it has proved possible to check Industry-standard design expectations by static tests on large offshore scale piles 2, 3, 4 . Research in several centres has emphasised the scientific weaknesses of the industry-standard American Petroleum Institute (API) RP2A 5 methodologies, which have remained practically unchanged between 1993 and 2007. While most practitioners have continued to use the conventional methods, alternative geotechnical design frameworks have been proposed that have been applied comprehensively in some sectors 6 . Vigorous debate has taken place over several years, prompted by industry-sponsored reports, academic papers, conference proceedings and meetings of the relevant API/International Organization for Standardization (ISO) review panels. Important changes are included in the 2007 API-RP2A recommendations for piles driven in sand that will also affect the ISO documents and industrial practice. However, progress is being made cautiously and further evolution of design practice can be expected. This paper offers one perspective on some of the issues raised in the recent debates, referring to background research and highlighting physical aspects of pile behaviour that are important to practice. Particular emphasis is placedon the question of axial capacity, as this is arguably the most important issue and has proved to be the focus for most discussion. Consideration is also given to the assessment of load displacement behaviour. While movement prediction and control is emphasised more strongly in onshore foundation projects, as reflected in the recent review by Mandolini et al. 7 , offshore engineers may become more concerned with making better fatigue life predictions for critical structures. Assessing limits for acceptable displacements also could become more important as a means of monitoring platform safety in critical cases. It is useful to consider at the outset the ranges of pile sizes specified by offshore engineers. Piles with diameters of ~5m have been driven for offshore wind turbines. Smaller diameters of 3 to 4m have been specified for such structures in the North Sea, where piles with diameters of up to 2.5m

ABSTRACT This paper deals with the design and installation of the foundation piles for the Skiff platform in the UK southern North Sea. The structure is novel because the well conductors are the primary platform foundation. The conductors were installed by drill-drive methods, for which conventional Soil Resistance to Driving)SRD) prediction method are not considered to apply. The paper shows that the Imperial College Pile design methods can be used to (1) obtain an SRD that gives a good prediction of blowcounts, and (2) provide a reliable estimate of the conductors? eventual static axial capacity. The paper compares predictions with actual installation data. INTRODUCTION The Skiff field is part of the Shell U.K. Limited (Shell) Sole Pit development in Block 48/20 of the UK sector, southern North Sea (Figure 1). The Skiff platform is an innovative slim-line design that was installed with a jack-up rig in April 2000 immediately before the wells were frilled (Figure 2). The jacket structure was placed on the seabed and tied back to the rig whiles the piled foundations were installed, and the topsides were place prior to drilling the wells. The main foundation elements for the platform are the well conductor, installed through the jacket legs. The conductors were installed using drill-drive techniques and axial pile capacity was based on the Imperial College Pile(ICP) 1 methods rather than the current American Petroleum Institute (API) RP2A recommendation 2 . As this was the first foundation of this type to be installed by Shell, extensive pile monitoring was carried out during installation and selected blowcounts were analysed by Heerema using the TNOWAVE computer program to match the driving stress waves. Figure 1: Vicinity map(available in full paper) Soil Conditions A geophysical survey carried out at the location revealed the water depth to be 26m and the soil profile to consist of some 75m of Holocene and Quaternary soils over Tertiary mudstone. Site investigation boreholes were drilled at the location to a maximum penetration of 78.2m below the seafloor. The boreholes confirmed the predominance of granular soils in the top 75m except for two minor clay layers above 11m penetration and a 3m thick layer below 31 penetration. A generalised soil stratigraphy and related design soil parameters, based on the site investigation data, are given in Tables 1. The design cone penetration point resistance q c? profile is show in Figure 3. Over-consolidation ratios (OCR) for the clay layers were found to be between five and ten, and this information was used to determine sand relative density, D r? using the procedures of Lunne and Christoffersen 3 . D r was assessed to vary from 80 to 65% with depth. Foundation Design The platform's plan dimensions were dictated by the specification that is must be installed by a jack-up rig with the piles driven with an impact hammer hung from the jack-up rig's draw-woks. This meant all the piles had to be reached by skidding the drilling contilever without the need to reposition the whole rig. The pile diameter was fixed to that of a standard well conductor

ABSTRACT This paper documents the use of the Imperial College Pile (ICP) design methods to obtain axial capacity of driven steel pipe piles for nine platforms in the UK North Sea. The diameter, wall thickness and length of the installed piles are based on these methods. The ICP methods have been used in preference to those given by API because they give more reliable values. The paper summarises site soil conditions, compares API and ICP capacities, discusses pile instrumentation data from installations and provides guidance on suitable factors of safety for use with the ICP methods. INTRODUCTION The pile capacity methods recommended by API have been successfully used to design foundations for offshore structures since the 1960s. The methods have recognised shortcomings and, in order to address these, the methods have been continuously improved and updated with the changes being documented in successive revisions to the published recommended practice, a process that continues to this day. In the early 1980s Shell U.K. Exploration and Production (Shell Expro) identified a need to improve design methods for its offshore structures, especially those relating to pile capacity in dense sand. To that end Shell Expro commissioned a major review of end bearing, for which a summary has been published 1 , and commenced sponsorship of a joint industry research project at Imperial College into shaft capacity that lasted ten years and culminated in the publication of a design guide 2 . As a result of this work Shell U.K. Limited (Shell) has not used API RP2A 3 pile design methods for platform foundations since 1996, when the Imperial College Pile (ICP) design methods 2 were first published. The reason for the change in design practice was because the new methods are based on improved soil models that give better predictability of pile load test results and they have been incorporated into the Company's Standards. The new methods allow Shell to demonstrate the safety of its designs through structural reliability assessments. In certain situations the new methods also result in smaller, more cost effective foundations. This paper summarises the soil conditions and foundations for nine North Sea platforms. It compares axial capacities derived by ICP 4 and API RP2A 3 methods with results from pile instrumentation that monitored pile-driving stresses during foundation installation. A procedure to assess foundation reliability is also described and target reliabilities are suggested. The results of reliability assessments for the platforms are given and the minimum resistance factors needed to meet the stated targets are tabulated. Shell has also used the ICP design methods to reanalyse the foundations of a number of existing platforms when updating structural assessments or considering additional topside loads, however, these platforms are not covered in this paper. The Sites Seven of the platforms are installed in the UK Southern North Sea (SNS), with the other two being installed in the UK Central North Sea (CNS). The platforms vary from 5 to 30m in plan dimensions and stand in 25 to 120m of water. They are all founded on driven steel pipe piles with diameters

ABSTRACT Numerous offshore wind farms in the German parts of the North and Baltic Sea with wind energy converters of a rated power up to 5.0MW have been approved up to now. Their special features are distances to shore of about 30km and more, water depths between 25 and 40m and a large number of wind energy converters up to several hundred plants per wind farm. These challenging conditions require an economic, sustainable and reliable foundation design. In general the choice of the foundation type depends on the loading characteristics and on the soil mechanical properties at the specific location. However, very often a monopile foundation has been found to be the best design solution. In the present paper, different design aspects of monopile foundations for offshore wind energy converters are evaluated. Special attention is paid on the characteristics of the monopile-soil interaction under cyclic lateral loading. A design concept is presented in which the results of cyclic triaxial element tests are combined with numerical calculations to predict the monopile behaviour. Development of German Offshore Wind Energy The German policy forces the extension of renewable energies to a level of 20% of the whole energy consumption by 2020. For this reason the currently installed wind power has to be expanded by utilisation of offshore wind energy resources. Due to a lot of restrictions in the German North and Baltic Sea, like existing nature reserves or shipping routes, most of the planned wind farms are located in the Exclusive Economic Zone in a distance to shore of 30km and more, where water depths range between 25 and 40m. Up to several hundred wind energy converters shall be installed per wind farm. For an economical operation, converters are needed with a rated power up to 5.0MW, which are still in test phase. Hence, major economic and technical problems exist and have to be solved, for example the grid connection, the maintenance and monitoring, but also the foundation design of the wind energy converters. Among the various foundation concepts, monopiles at present are a clear and consequential solution to carry the structural loads transferred by the tower of the wind energy converter into the subsoil 1 . Monopile Behaviour under Cyclic Lateral Loading Load characteristics The loading of a monopile foundation results from wind, waves and currents. These loads are highly stochastic in nature. Although their effects on the structure can be more or less reliably estimated, it is not easy to quantify and assess the loading which is relevant for the foundation design. Especially the reduction of the complex loading to a series of simple cyclic load parcels, i.e. load collectives, is difficult. Regarding the wave loads it is relatively clear how to model a cyclic loading sequence characterised by wave height, wave period and ideally also wave direction. Available data exist which allow the configuration of a load collective, e.g. Hapel 2 . For this reason only wave loads are considered in the following. The transformation of wind and currents to load collectives requires more research efforts.

ABSTRACT As operator of Block 17 offshore Angola, Total E&P Angola has developed three oil fields, all in about 1300m of water and with uniform soil conditions made of soft highly plastic clays. With Sites A and B floating production storage and off-loading (FPSO) piles and Sites A and C riser tower foundations having similar dimensions, the comparison of their installation behaviour is particularly relevant for highlighting some specific features. The Site A installation results were significantly below predictions, because of (a) reduced friction along partly painted steel piles and (b) overestimation of the penetration resistance for the first suction piles ever installed in such West Africa deepwater clays. The installation results for Sites B and C were in agreement with the predictions, thus confirming the validity of the penetration analyses. INTRODUCTION As operator of Block 17 offshore Angola, Total E&P Angola has developed three oil fields in about 1300m of water (named Sites A, B and C), all located in the same area of Block 17, which represents a 90km x 60km rectangle with a bathymetry ranging between approximately 300 and 1900m. Both Site A and Site B developments are based on subsea production systems, associated with a large FPSO unit and an export buoy, respectively with 40 and 70 wells spread over an area of about 50km 2 . The Site A subsea wells and flowlines are linked to the FPSO by means of three riser bundle towers, while flexible catenary risers are utilised for Site B. The Site C development includes about 20 wells and the produced oil is sent by flowlines to the Site A FPSO via a fourth riser tower. The distance between the Site A FPSO and the Site B and Site C fields is approximately 10 and 15km, respectively. Suction piles are used to anchor both FPSO vessels (Figures 1 and 2) and export buoys, and suction caissons are utilised as foundations for the four riser towers (Figures 3 and 4) as well as for all subsea wells manifolds 1 . The geotechnical design of the suction piles for both FPSO vessels and of the caisson foundations for the riser towers was carried out by the Norwegian Geotechnical Institute (NGI), and the moorings of the two FPSO vessels were certified by Bureau Veritas (BV). Soil Conditions and Geotechnical Design Profiles In the central part of Block 17, the seabed slopes gently to the south west with a gradient of about 1:50 (1.2°). On the basis of high resolution, 3D seismic data, the thickness of the clayey sedimentary infill is estimated to be equal to approximately 300m. From about 40m below the sea bottom, numerous small faults or fractures are present, some of them probably extending to the seabed. Locally, specific features are encountered, such as salt diapirs, pockmarks, cold seeps or carbonate concretions. The soil conditions, being consistent and relatively uniform, are composed of high plasticity clays, characterised by a plasticity index, I p' , over 100%, a low effective unit weight

ABSTRACT Skirted, or bucket, foundations are widely used offshore to support or anchor structures for the oil and gas industry and provide an attractive foundation solution for future developments. Current industry design guidelines give conservative collapse loads for offshore shallow foundation systems and loading conditions, particularly due to overlooking the beneficial effect of negative excess pore pressure developed within the soil plug during rapid uplift or overturning. This paper outlines research being carried out at the Centre for Offsore Foundation Systems (COFS) at the University of Western Australia to investigate the response of skirted shallow foundations, to develop reliable enhanced predictions of limit states under transient loading, and to quantify the timescale over which sustained tensile loading can be withstood. The project involves using centrifuge modeling to investigate the effect on capacity of skirt dept; consolidation stress level; time after installation when uplift is applied; eccentric and lateral loading in conjuction with uplift; conditions in which gapping along the foundation skirt occur and their effect on uplift capacity; and methods of suppressing crack formation. Background Skirted foundations have application for a variety of offshore structures, as shown in Figure 1. They can be embedded to depth of up two foundation diameters, although shallower embedment ratios are often associated with larger foundation diameter. Circumferential skirts confine a soil plug, allowing uplift resistance to be mobilized by suctions (i.e. negative excess pore pressures) developed within the soil plug. This is particularly relevant offshore, where environmental forces lead to large horizontal and moment components of foundation loading. In general there is a lack of accurate solutions for general loading of skirted foundation, and there is uncertainty over, and no formal guidance regarding, the timescale for which tensile stresses can be sustained beneath a skirted foundation. Routine offshore design guidance for shallow foundation 1,2,3 is based on classical bearing capacity theory 4,5 , which has several shortcomings in respect to offshore foundation systems and loading conditions. Formal design guidance does not represent the simultaneous action of horizontal load and moment 6,7,8,9 , uses conservative, empirical depth factors 10,11,12 and neglect tensile capacity. The size of offshore shallow foundation (up to 15 000m 2 for a gravity-based structure) precludes direct assessment of capacity by mean of loading tests, and reliance must be placed instead on results from numerical analyses and physical model tests. The latter has been carried out under laboratory conditions but without the correct scaling of soil strength and self-weight stresses 13,14,15 . This is a significant limitation for problems involving uplift, or where potential cracks or gaps may develop between structure and soil. Model tests conducted in a geotechnical centrifuge overcome this problem, but existing centrifuge data are limited in terms of soil conditions, foundation geometry and applied load combinations 16, 17, 18 . Greater progress has been made in numerical modeling 8,9,19,20 . However, the main body of work is restricted to undrained soil response, plane strain geometry and surface footings with full bonding at the soil interface to represent tension capacity.