Ursa is the largest TLP in the Gulf of Mexico having 98.000 tons displacement and is the deepest TLP in the world located in 3,800 feet-of-water, MC Block 809. The URSA tendon and foundation design, fabrication and installation were built on the wealth of experience of previously installed TLPs1,2,3,11. In this paper, we focus on the new challenges that had to be met for this record setting project. Unique features of the Ursa Tendons, Foundation and Installation include:More tendons, 16 (4/corner) versus 12 (3/corner);Larger tendons, 32"OD × 1.5" wall (compared to the previous 26" × 1.3" and 28" × 1.2" wall)Longer Tendon Main Body Sections than the previous TLP's, resulting in 283.5' individual sections. This reduced the number of connections required and offshore handling time.New tendon top connector devdoped for Ursa resulting in simpler operations and significant cost savings.The foundation system consists of 16 individual piles. They are 417' in length and 96" OD constructed from 60-ksi material.The potential for significant fatigue-damage accumulation due to pile driving required special design and fabrication solutions.Design, Construction, and Installation strategy developed early on helped reduce cycle time, cost, and accommodated changes during construction.On site temporary mooring of the TLP was accomplished using a Flex-Yoke system.Improved procedures for tendon handling, stabbing and tensioning were developed. Introduction The Ursa TLP tendon, foundation and installation design and fabrication extended over a period of 3 years. In order to meet the fast-tract project schedule, material orders had to be placed early, before design completion. Design decisions were focused on reducing both cost and schedule, taking advantage of learnings from the previous TLP's, Early attention was given to the TLP installation design to allow modifications to be incorporated into the planned schedule, tendon design and pile driving requirements. The modification brought out by the site and water depth change6,7,9 were significant. New soil information had to be obtained, designs verified, installation procedures adjusted and schedules maintained. Tendon Design, Fabrication and Transportation TLP and Tendon Sizing After concept selection in the summer of 1995, the TLPSIZE program was used to size and optimize the URSA TLP. For the initial site, the water depth was determined to be 3,950 feet. Later, due to the site relocation, the water depth was reduced to 3,S00 feet7. Cost optimization was focused on sizing the hull and tendons such that the minimum tendon bottom tension was slightly above zero and the tendon met stress, collapse, and fatigue design criteria with minimal steel cross section area and maximized buoyancy. The program was calibrated based on technical and cost data from the MARS TLP.

No abstract

A discrete-element method of dynamic beam columns, which are rest.ing on linear or nonlinear-elastic, or nonlinear-inelastic supported and subjected to either static fixed loads or dynamic loads, is presented. Two parallel series multielement models, which consist of a collection of elastic and slip elements, are used to simulate the inelastic characteristics of supports. Fracture or softening of the supports is not. considered, but linear viscosity can be included. Potential applications include study of the dynamic response of offshore piles to wave-induced forces and earthquake-induced forces and prediction of the hysteretic effect of inelastic supports under pavement slabs. INTRODUCTION In recent years, a great deal of interest has been focused on the utilization of numerical methods to describe computer models or problems in struct.ural dynamics. This work has been devoted to the development of a beamcolumn computer program which uses multielement models to predict the loading paths of resistance-deflection curves of the nonlinearinelastic supports. The multielement models for the nonlinearinelastic supports described in this work are not time dependent; therefore, relaxation (creep) of resistance of the support is not included in the model. The retardation, or delayed elasticity of' the support can be considered by installing an external viscous dashpot in parallel with the support m6del. Fracture or softening of the nonlinear-inelastic support are also not included but strain hardening may be considered. Crank and Nicolson2 have introduced an implicit formula to solve second-order heat flow problems. Salani7 is credited with applying this implicit formula in determining the transverse time-dependent linear deflections of a beam or plate. Essentially, the beam is replaced by an arbitrary number of rigid bars and deformabie' joints, and time is divided into discrete, equal intervals. The representation readily permits the flexural stiffness the elastic restraints, the mass densities, and the applied external loadings to be discontinuous and lumped at the deformable joints that connect the rigid bars. The governing partial differential equation at each joint is approximated by a difference equation that includes several unknown deflections of the joint, whichoccur at specified time intervals. All difference equations are based on the assumption of linear elasticity and elementary beam theory. The nonlinear characteristics of the supports are considered by using either a spring-load iteration process (adjusting both the stiffness and the load from one iteration to the next, which is known as tangent modulus method) or a load iteration technique (adjusting only the load). The iteration process compares successively computed deflections until a specified tolerance is satisfied. Only three nonlinear characteristics of support curves are considered. The first is exhibiting the same resistance to either upward or downward deflections, hereafter referred to as symmetric resistance-deflection curve.

During the years 1981 through 1986, Conoco, Inc. sponsored a comprehensive program of research aimed at improving the understanding of pile-soilinteraction along driven piles subjected to static and cyclic tension loading. The goal of the research was the development of guidelines and procedures for the design and analysis of foundation piles for Tension Leg Platforms. The experimental work was done at a decommissioned CAGC platform at Block 58A of the West Delta Area of the Gulf- of Mexico. Included were tests on a fully instrumented 30-inch-diameter pile driven to a penetration of 234 f t and eighteen experiments with instrumented probes. This paper contains the results of the experiments with the instrumented probes, with comparisons made to similar experiments at other sites. INTRODUCTION In 1981, Conoco, Inc. initiated a five-year study of axial pile-soil interaction along driven piles subjected to static and cyclic loading. The study was performed by The Earth Technology Corporation, acting as a designated subcontractor to Conoco Norway, Inc. through A.S. Veritec. The work included field experiments observing the behavior of fully-instrumented piles and pile models, along with parallel analytical work aimed at increasing the understanding of the fundamental behavior of axial pile-soil interaction in normally consolidated clays, with emphasis on applications for Tension Leg Platforms. In order to ensure a high degree of confidence in the extrapolation of the results to a site in Green Canyon Block 184, a search was made for a site having homogeneous soft-to-stiff normally consolidated clays with engineering properties very similar to the deepwater site. Such a site was located in Block 58 of the West Delta Area of the Gulf of Mexico. A decommissioned production platform located at the site, West Delta 58A, was refurbished for use as a testing platform. The major part of the work involved static and cyclic load tests on a fully instrumented 30-inch-diameter pile which was driven to a penetration of 234 ft below the seafloor. Load tests were performed at times ranging from 2 hours to 29 months after driving, times which encompassed the complete consolidation history of the near-pile clay. Early in the work, it was considered to be extremely important to collect parallel data from in-situ experiments using small-diameter probes for comparison with the results of the large-diameter pile tests. The use of small-diameter probes allowed a much wider variety of test conditions to be employed than was possible with a single large-diameter pile, particularly including variations of consolidation time and loading history. Two types of instrumented probes were employed, one 3.00 inches in diameter and fitted with an opened cutting shoe and the other 1.72 inches in diameter that had a solid conical point. The probes were installed beyond the bottoms of boreholes and provided continuous measurement of pore pressure, total lateral pressure, and axial shear resistance, all as a function of axial displacement. In the constant temperature environment of deep soil embedment, the instruments exhibited excellent stability and repeatability. They have been described in detail previously (Bogard, et a1 1985).

No abstract