Abstract This paper describes the design, manufacture, and installation of subseaumbilicals, risers and flowlines (SURF) interface on the OPTI-EX floatingproduction system (FPS). For the purposes of this paper, the SURF interfacewill be limited to the riserto- hull structural connection. This structuralconnection consists of the riser end fitting, riser porch, and riser porchadapter. The riser porches are very efficiently designed. For the OPTI-EX, 7.5MT of steel weight has a top tension capacity of 300 MT. The design is highlyflexible at accommodating field architecture, required no internal hullmodifications or foundations, and the OPTI-EX, located on the WHO DAT field, accommodates flexible riser sizes from 5 inches (for production risers) to 11.5inches (for oil export lines) in 1000 m water depths. The riser porches and mating riser adapters/I-tubes are simple structures;precision manufactured to ensure proper load distribution in-service and toensure ease of in-yard and future offshore installation. The design is robust, fatigue resistant, and fabrication-friendly. Mounted on the outboard facing surfaces of the FPS hull pontoon, these mountsfacilitate riser installation and pipeline laying in directions away from theunit. At WHO DAT, the riser adapters accommodate pipeline approach anglesrelative to the FPS from perpendicular to a given side of the unit to as muchas 26 degrees from perpendicular thus providing great pipeline layoutflexibility relative to the unit. The rigging of the flexible pull-ins, though " involved", was not overlycomplex. Maximum pull-in loads were low as the catenary span of all risers atinitial hang off was short. This short length was due to utilizing the " layaway" method of deploying the flexibles once hung off. This allowed pull-inequipment on the FPS to be relatively small and relatively easy to handleoffshore. Introduction OPTI-EX was speculatively built as a generic FPS able to operate in a broadrange of water depths and MetOcean conditions. Since it had no operator andrelated specific field location, field architecture and riser design specificswere unknown prior to unit fabrication. Hence, no riser porches weremanufactured and installed during construction even though a riser porch designhad been developed for it that would accommodate a wide array of riser sizes, weights, and configurations utilizing either flexible risers or SCRs (steelcatenary risers). The porch design was such that the unit could accommodate upto 20 risers. The maximum single riser design top tension capacity was 500 MT. This design was presented to LLOG and its pipeline engineer for use on the WHODAT project. It was selected as the solution to accommodate up to 12 productionflowlines and 2 export pipelines. Only three production risers and two exportrisers were initially installed, the rest of the porch slots were reserved forfuture production risers. A fourth production riser will be installed duringthe first quarter of 2012. When installed, the four production risers willcomprise two each piggable flowline loops, one loop serving one well cluster. Refer to Figures 1 and 2 for field layout.

Abstract The paper describes the design, installation and operation of a Ship to Ship (STS) transfer system for LNG. It addresses the qualification of the cryogenic hoses and the emergency-disconnect couplings as well as the operational procedures for the berthing and transfer operation offshore. The system uses ‘standard’ components, adapted only where needed for its new function. The operational procedures for berthing offshore have a long and proven track record in the oil lightering industry. The net result is a simple and cost effective solution for offshore transfer of LNG, developed in a fast and efficient manner and already proven with more than 10 commercial operations.

Abstract The author was the client representative on an EPC contract between Triton Energy (now Amerada Hess) and Stolt Offshore for flowlines and risers in a deepwater West Africa development - the Ceiba field. The following are some lessons learned during the survey, design, installation, testing and commissioning of the flowlines and risers from the subsea manifolds in water depths of 650 - 750 m. to the FPSO in 90 m. water depth. Comparisons will be made between a surface vessel deployed SWATH multi-beam survey and a ROV mounted multi-beam bathymetric survey (of the seafloor's XYZ coordinates) - and their resulting seafloor profiles. The predicted pipeline spanning will becompared with the actual measured span profiles. The evolution of a mechanical pipeline support design from inception to successful deployment will be reviewed, along with an unsuccessful design. The span support criteria selected for this field will be explained in light of the geotechnical survey data and bathymetry. Introduction General The bathymetry of the field, established by a SWATH survey and shown in Figure 1, has deep meandering channeling with numerous deep depressions (a crosssection of one of the channels is shown in Figure 2). The channels point in the general direction of the mouth of the Rio Muni River, separating Equatorial Guinea from Gabon. The SWATH survey database was used to create the flowline profiles for the spanning analysis. AGA Level 2 (ref. 1) was used for stability calculations. Field Architecture The field architecture has subsea wellheads and manifolds in 650 to 750 meters of water. Dual flowlines connect the FPSO to the manifolds. Flexible pipe jumpers/CVC (Cameron Vertical Connector) connectors connect the rigid flowlines to the manifold. At the FPSO, flexible pipe risers (which are routed over a riser support structure (RSS) and up to the FPSO - forming a Lazy S profile), connect the rigid flowlines to the FPSO. The requirement to have inhibitor within 48 hours of flooding the flexible pipe with seawater resulted in the necessity of a wet lift to install the jumpers and make the CVC connection to the manifold. This meant that span remediation for hydrotest conditions would have to be completed early in order to maintain the construction schedule. This requirement ensured active participation from the installation vessel crew in the span support design process. Soils Investigations Figure 3 shows the drop core locations for the soil sampling, with the Stations 1, 5 and 39 drop cores that were tested at a soils laboratory (ref. 2). The soil testing included the Atterberg Limits (Liquid Limit, etc.), thermal conductivity, electrical resistivity and particle size analysis. Vane shear tests were also performed to obtain the shear strength at the liquid limit. In addition, relative density testing was performed on the granular material at the FPSO (for AGA, Level 2 Stability calculations). For Stations 1, 5 and 39, respectively, the ratios of the mudline soil water content to the Liquid Limit were 0.84, 0.79 and 1.19, with undrained shear strength values of 8, 6 and 4 kPa. The seafloor soils are turbidite deposits that range from sandy at the FPSO to silty clays at the wellhead locations.

ABSTRACT On the British Gas North Morecambe Development Project in 1993, the derrick-lay barge DLB 1601 used a dual lay operation. A 36" Ø pipeline was laid piggybacked onto a 36" Ø pipeline for the shore pull operation and then separated to follow a dual lay for the offshore section. Transformation for piggyback laying mode to the dual /ay mode was performed making use of a new/y developed transition operation. This paper reports the challenges which had to be met to ensure a successful outcome. Results of various engineering studies are first presented. The set-up and observations of the (onshore) prototype testing are then described, Lastly, the special measures taken for the transition operation and certain field observations during the installation phase are briefly discussed. It is beiieved that the information presented in this paper will be of use to other similar projects in the future. INTRODUCTION For the offshore oil and gas industry, thousands of kilometres of submarine pipeline have been constructed all over the world. With the conventional S-lay method, the single-pipe laying technique and the piggyback (or bund[e) pipelaying technique are quite well known. While the former technique involves the laying of a single pipeline, the latter allows eventually the bundle laying of two or more pipelines. In the piggyback pipelaying technique, the small pipelines are generally attached to a comparatively larger pipeline which passes through the tensioners. Two equal-size pipelines have also been laid as a bundle, where both lines were passed through the tensioners. In this technique, different pipelines are fabricated in parallel on the lay vessel. Using saddles and straps, these pipelines are attached to form a bundle near the vessel stern. This thereby allows the pipelines to leave the lay vessel as a bundle and facilitates to lay all the pipelines together on the seabed. In such a case, therefore, touch-down point monitoring of the bundle pipeline is of importance similar to the single-pipe laying technique. Simultaneous laying of two (dual lay) or more (multiple lay) unattached pipelines from the same pipelay vessel are indeed scanty. This pipelay technique is only used in special cases where the pipelines are to be constructed very close to each other, keeping a specified separation. This may be due to any requirement, such as constructional constraints, protection, maintenance, etc. As a result, each pipeline has to leave the lay vessel separately. It necessitates the need for individuallaying ramps having tensioning arrangements for each pipeline. Lay vessel preparation and monitoring of the overall operation associated with such a laying technique therefore need special attention. Over the last two decades, the derrick-lay barge DLB 1601 has performed a number of challenging pipelaying projects. Among these projects, the British Gas North Morecambe Development (NMD) project in 1993 needs a special mention. This is particularly due to the fact that DLB 1601 performed a dual lay operation for the first time, Moreover, an interesting transition operation from piggyback to dual lay was also carried out. This paper describes the engineering and installation challenge.

ABSTRACT A test program consisting of 13 steel tubular brace specimens with dent damage and various diameter-to-thickness (D/t) ratio was conducted to assess the residual strength and repair of damaged tubular braces in offshore platforms. Specimens were inflicted with a dent having a depth of 0.1 OD, with internal full grouting and grouted steel clamps, respectively, used to repair specimens. A baseline of unrepaired, damaged specimens subjected to concentric axial loading showed a significant reduction in residual strength due to the dent. A further reduction was found in specimens subjected to combined loading. The use of grout to repair specimens was found to inhibit a growth in dent depth under axial loading and restore the capacity of the dented member to a level greater than the undamaged strength. A comparison with existing and modified analytical methods for strength prediction of unrepaired and grout repaired braces shows these methods to provide reasonable results. INTRODUCTION At present time there are more than 6,000 offshore platforms worldwide. Most of them were designed for a 20 year life but are still in operation after 30 to 40 years after installation. Many platforms have suffered some form of damage. Damage records indicate that the majority of accidents result from either ship collisions or impact with debris, including dropped objects. These accidents account for 177. and 12?40, respectively, of all recorded accidents [1]. Under impact, the memberâ??s circular cross section is susceptible to localized denting and ovalization. Previous investigations on dented members have shown that even minor damage can cause severe degradation in a memberâ??s strength, which in turn affects platform integrity. Economic necessity and the need to ensure structural integrity and safety creates strong incentives to develop techniques to rehabilitate damaged offshore structures. Previous studies on the residual strength include both experimental and analytical research. The experimental studies reported in the literature include the work of Smith et al. [2,3], Taby [4], as well as Landet and Lotsberg [5]. Analytical work on dented members includes the nonlinear finite element studies conducted by Maclntyre and Birkemoe [6], the elastic-plastic beam-column analysis model of Smith [2,3], as well as a moment-curvature approach to predict nonlinear behavior by Duan et al. [7] and Padula and Ostapenko [8]. Based on experimental data, there have been computer programs and closed form solutions developed to predict the residual strength of dented members. One is that of Taby [9], who developed a semi-empirical method for predicting dented member behavior using a stress resultant approach applied to a beam element. The method was implemented into a computer program called DENTA. Ellinas [1 O] proposed a quadratic equation for ultimate strength prediction based on equilibrium of concentrically axially loaded dented members. In addition to studies of dented members, some investigations have been conducted on grout repaired members having dent damage. Boswell et al. [11] has conducted a series of mostly small-scale tests on complete grout filled dented braces as part of a United Kingdom Department of Energy project, as well as tests by Parsanejad [12].

ABSTRACT This paper presents the background to the design of eight stressed grouted clamps used to alleviate fatigue problems associated with T and X joints on two platforms in the Gulf of Mexico. Offshore design codes do not cater for the design and specification of clamp systems. The paper therefore addresses the design philosophy and approach adopted, together with the specifications and procedures developed during the design stage. The clamp technology discussed in this paper represents the first major application of such systems in the Gulf of Mexico, and follows on from their extensive use for platforms in the North Sea. The need to strengthen or repair jacket structures has been increasing over recent years, and the clamp technology described is recognised as cost-effective, since it offers an alternative to the use of underwater welding techniques. The design of the clamp system is reported, covering both the static and fatigue aspects. Procedures related to sealing, grouting and bolt stressing are described with specific reference to the Gulf of Mexico environment. Clamp systems are expected to find increasing favour in the strengthening and repair of existing installations in the Gulf of Mexico and elsewhere. The systems described in this paper have been designed and installed on two Gulf of Mexico platforms, and it is demonstrated that such systems are both technically efficient and cost-effective. Clamp systems are extremely versatile, and can be adapted to meet the strengthening/repair needs for all forms of damage, eg. fatigue cracking, corrosion, supply vessel collision, dropped objects, static overload or design omission. INTRODUCTION The repair of existing offshore installations is now an important part of offshore engineering. Design codes provide little or no guidance in this area. The need to repair or strengthen offshore installations is likely to increase as the number of installations increase and as the existing structures get older. In addition, this need becomes more likely as shipping movements in the oil and gas fields increase and as more becomes known about the environment in which installations operate. There are anumber of causes of damage or defect which may require repair, and include:- dropped objects installation damage vessel collision crack-inducing fatigue loads design upgrade corrosion welding and/or fabrication fault. This paper is concerned with the repair and strengthening of two Gulf of Mexico platforms, using clamping technology. The paper has been divided in two parts:- A - Background B - Design philosophy, specifications, and procedures The design, fabrication and installation of the clamps were carried out in 1990. This paper concentrates on the design aspects. PART A - BACKGROUND Damage Appraisal The two Gulf of Mexico platforms are referred to in this paper as Platforms 1 and 2. A part plan at El (-)120' is shown in Figure 1, and a part plan at El (-) 225' appears in Figure 2. Both platforms are similar in configuration, and the plan arrangements shown can be considered to be typical of both platforms.

ABSTRACT This paper describes the installation of the Brae 'B' platform in the UK sector of the North Sea. Over 72,000 tonnes of facilities were installed within a period of 67 days in the early summer of 1987. The work included launching a 22,500 tonne jacket, driving 35 piles, depositing 4.5 tonnes of weld metal and lifting 35 modules. Also, a small tripod structure was set and two 50m pipeline spools were successfully connected to the existing pipelines. INTRODUCTION Marathon Oil (UK) operates the Brae Field as a joint venture with the following companies: Britoil plc Bow Valley Exploration (UK) Ltd Kerr-McGee Oil (UK) Ltd Dyas Oil UK LL&E (UK) Inc Sovereign Oil and Gas Ltd Norsk Hydro Oil and Gas Ltd Brae 'B' is the second platform to be installed in the field which is located 155 miles north-east of Aberdeen. The first (Brae 'A') was installed in 1982 and was considered a large platform at the time with a topside design weight of 36,000 tonnes. The Brae 'B' platform, however, surpassed this with a total operating topside weight of 42,000 tonnes. The modules and jackets were fabricated at seven UK sites and a fleet of 16 cargo barges was required to move the components offshore. In all 38 major separate lifts took place to complete the platform which included an off-structure flare tower. Detailed planning was an important feature of the project. The installation itself was scheduled using a 300 activity probabilistic model which incorporated known weather sensitivity factors. After much fine-tuning the final run showed that the installation period could take up to 95 days. In fact it was accomplished in an overall duration of 67 days which included 18 days downtime. As a result, the platform hook-up was given an early start and first oil is now expected well ahead of the original schedule. OUTLINE DESIGN The hydrocarbon reserves in the Brae Field are a complex mix of gas, oil and condensates. The Northern platform, Brae 'B', is expected to be producing 75,000 bpd of condensate and 400 million cubic feet of wet gas by the end of 1990. Because much of this gas will be re-cycled initially, the largest ever offshore gas processing plant is now incorporated on the platform. This includes four Rolls-Royce RB211 gas turbines to handle the compression. A further three RB211 turbines are provided to give the platform 72MW of electrical power. In addition, two complete drilling rigs have been installed to cope with the demands of the Brae reservoirs and the drilling programme. Some of the other vital statistics of this substantial development are given in Figure 1. INSTALLATION DESIGN From the outset the installation techniques available were considered before selecting the final shape of the major components of the platform.

ABSTRACT Most of tubular joints in offshore structures have quite a number of braces and may be loaded in various ways. It is of great significance to find out the maximum stresses in the multi-brace non-coplanar tubular joints under complex load condition. Based on the results obtained from stress analysis of a set of tubular joints, the leading feature of stress distribution in the multi-brace non-coplanar tubular joints is discussed and a method for calculating the maximum stress is proposed. Furthermore, the calculating formulae are derived. INTRODUCTION Stress concentration factors of some simple tubular joints under typical load condition have been studied conscientiously. There are several empirical formulae available for the SCFs of T-, Y-, X-, K-and KT-joint under axial forces, in-plane moments and out-of-plane moments (1-4) . But the actual situation of tubular joints in offshore structures is far more complex. Firstly, it depends on the load condition at which point the maximum stress occurs. In T-joints, for example, the maximum stress usually occurs near the crowns under in-plane moments or at the saddles under axial forces and out-of-plane moments. Fig. 1 shows the movement of the maximum stress point as the moment changes in deriction. Obviously, when the maximum stress occurs at saddles, the stresses at orowns are not equal to zero and vice versa. The empirical formulae available are only for the value of the SCFs, they can not be used to specify the exact locations of the maximum stress and to describe the stress distribution near intersections of the chord and braces. In practice, a real load can barely be just an axial force, an in-plane moment or an out-of- plane moment. The maximum stress can occur at nearly any point along intersections. stresses at different points can not be superimposed at all. Therefore, non of the formulae available is applicable in general1 load condition. Secondly, the SCF is the non-dimensional magnitude of the maximum principal stress and the principal axes of stress in different stress state do not coincide with each other. Fig. 1 shows that if deriction of the moment change from ?= .0° to ?= 90°, the principal axes of stress at the saddles rotate through an angle of 45°. Thus even the maximum stress point is predicated, its value can not be obtained by superimposing the principal stresses in different load condition either. Finally, the stress distribution will differ greatly from each other when many of the braces of a tubular joint are loaded in different ways. Reference (5) has pointed out that for the tubular joint shown in Fig. 2, the SCF in Case I is 2.4 times as big as that in Case III, but the magnitude of the forces exerted on the braces remains unchanged. In order to evaluate correctly the working life, the research work on calculation of the maximum stress in multi-brace non-coplanar tubular joints is necessary. Because of complexity of the geometrical configuration and load condition, the stress analysis of multi-brace tubular joints is difficult and costly. In our study, a computer program system, named 'FEAPS-TJ', has been developed for the mesh generation and finite element analysis.

ABSTRACT An alternative procedure for predicting the combined hot-spot stress (CHSS) at tubular K and Y joints under combined branch loading is presented. The procedure makes use of influence factor (IF) equations developed, as a function of the joint geometry and branch loading, for various potential hot-spot, at a point, the effects of the axial force and bending moments acting on each branch. The locations on the branch and chord sides of the weld. The CHSS is obtained by linearly superimposing resulting CHSS, therefore, reflects location, orientation and sign of each branch load contribution. Comparisons of predicted CHSS obtained via the new and other procedures to stresses from finite element analyses were made on a large sample of joints. Result show that (1) the new procedure is substantially more reliable than the other procedures studied, (2) none of the procedures consistently predicts conservative CHSS values, and (3) the overriding factor influencing the accuracy of the CHSS calculations appears to be the accuracy of the parametric equations. Although a better stress predictor can be expected to yield more reliable fatigue damage estimates, damage calculations will exhibit broad scatter due to the power function relating damage to stress. Unfortunately, further improvements in the accuracy of CHSS based on parametric equations are not likely to be easily achieved, given the large number of variables and locations that need to be considered. INTRODUCTION Tubular joints are extensively used in offshore structures, particularly in steel production platforms. A tubular joint consists of a through tubular member (chord) to which other members (branches) are we1ded. Because of the cyc1ic nature of the wave action on the structure, the joints are subjected to cyclic loads and possible fatigue damage. Depending on the water depth and the dynamic response of the structure, fatigue of the joints can become the controlling design factor. Among the various factors affecting the fatigue analysis of tubular joints, the combined hot-spot stress (CHSS) plays an important role. CHSS is commonly associated with the maximum stress or hot-spot stress occurring at the branch chord intersection of the joint when the branches are subjected to the combined action of axial forces and in-plane and out-of-plane bending moments. To place the CHSS within the context of the fatigue design procedure, the load and resistance sides of the design should be considered. Figure 1 gives a simplified flow chart of the approach to the load and resistance sides of the design. The resistance side involves the development of S-N curves through experimental testing. On the load side, the CHSS history caused by the wave-structure interaction must be determined. The fatigue damage is estimated as the ratio of the load to the resistance, both expressed in terms of cycles (P-M rule assumption). The resulting damage is proportional to a power function of the stress and, as such, small changes in stress result in large changes of estimated damage or life. Hence, proper assessment of the CHSS value is important.

ABSTRACT This paper describes the concept, analysis, design, and sea trials of an active/passive ten ton capacity motion compensating crane for deploying a remote unmaned work system. The same concepts can be employed to make possible the handling of tethered loads, including ship to ship transfers, in high sea states without the risk of cable failure or excess payload pendulation. Motion compensation is achieved by driving the boom up and down, while the ship is heaving, such that the boom tip remains substantially stationary with respect to a fixed point on earth. The combination of active and passive boom control resulted in significant savings in power consumption over a purely active system while providing excellent motion compensation. The crane; mounted at the stern of a salvage tug, has a reach of 30 feet (9 m), a boom tip stroke of 24 feet (7.3 m), and can deploy the vehicle over the side or stern. A traveling saddle on the boom positively restrains the payload during deck handling operations while moving the load inboard or outboard. The cable storage reel accommodates over 23,000 feet (7000 m) of Kevlar wound electromechanical cable, and the entire crane can be disassembled into modules for air transportability. Performance data is provided for a sea state four operation, based on computer simulation predictions. INTRODUCTION The safety and performance of offshore operations are severely limited by wave-induced vessel motions, and as a result work is often suspended during heavy weather. Curtailment of operation can be very costly, and possibly intolerable, as in the case military operations. In 1973, a requirement for a motion compensating crane was identified for launching, recovering and 6. handling a tethered underwater work system to 20,000 feet (6000 m) depth. Analysis using digital computer simulations indicated that the existing technology of passive systems would not meet the performance criteria. Therefore, the development of an active system was initiated, with the prototype being completed and ready for sea trials by early 1977. SYSTEM REQUIREMENTS The general arrangement of the remote unmanned work system is shown in Figure 1. Our basic requirement was to provide the "Motion Compensation Rotating Gantry Crane" which enables the work system to operate successfully in a Sea State 4. The key specifications for the cable and payload are given in Table 1. In addition to being extremely long, the Kevlar cable is very fragile, having little resistance to abrasion, flexure and fatique. The crane, therefore, must treat the cable "gently" in all facets of operation: deck handling, launch and recovery, and deep submergence. As a result, the following primary requirements have been established. The cable must be stored at low tension. This necessitates a separate traction winch and storage reel. All bends must be large diameter, and inside sheave surfaces must be soft and smooth. Cyclic bending of the cable for motion compensation cannot be tolerated; this rules out a "constant tension" winch in favor of an oscillating boom.