The search for offshore petroleum reserves has taken the energy industry into deeper waters and deeper wells. A byproduct of deep-water discoveries is often a high-pressure, high-temperature (HPHT) reservoir. Reservoir fluids are often transported to a floating production facility (FPF) via carbon steel flowlines and risers. Pressurization and heating of flowlines that were installed empty and cold can lead to significant longitudinal expansion.If left uncontrolled, significant expansion can often lead to excessive stresses and strains in flowlines or other system components. In some cases, uncontrolled expansion can lead to local buckling of the pipe cross-section or high stress-range (low-cycle) fatigue failure, resulting in a leak to the environment.The Tahiti Project in the Gulf of Mexico presented a number of different flowline expansion system problems, including high pressure, high temperature, future water injection and reservoir souring, global lateral buckling and fatigue associated with a number of operating cycles, high flowline end expansions, high riser bottom tensions, and flowline walking (soil ratcheting). Additionally, the stringent design challenges of the project produced corresponding challenges in the area of welding.This paper discusses some of the flowline expansion control techniques used on the Chevron Tahiti Project in the Gulf of Mexico. Midline expansion control via distributed buoyancy and end expansion control via a novel stab and hinge-over (patent pending) Pipeline End Termination (PLET) located on top of a suction pile are discussed, along with other expansion control topics. The welding requirements, the challenge of setting up an onshore multi-joint facility, welding strategies, and the corresponding results will also be discussed.

Design of Steel Catenary Risers (SCR) has seen increasing challenges in deep water applications due to higher pressures, vessels with more dynamic motions and severe weather conditions. In the Gulf of Mexico, the SCR is sometimes designed with little margin due to harsh environment. In West Africa, SCRs may be infeasible with turret moored Floating Production Storage and Offloading (FPSO) System, and are often limited to a narrow hang-off range near the Center of Gravity (CoG) with spread moored FPSO. Freestanding hybrid risers are feasible solutions for many fields in West Africa. However, the freestanding riser cost is generally considered much higher than that of an SCR. Hence, there is a need to improve SCR performance in dynamic vessel applications. This paper presents the development of Shaped SCR configuration that has significantly improved performance and minimal cost increase compared to an SCR. Compared with a conventional Lazy Wave SCR, the riser configuration has increased installation flexibility and decreased cost as it is installable via J-lay or S-lay. The configuration is developed and assessed for a spread moored FPSO in West Africa. The strength and fatigue response is compared with those of a simple SCR and Lazy Wave SCR. The key considerations in developing the configurations are discussed. This paper also presents the use of upset ends to improve riser fatigue performance. This is achieved due to reduced bending and tension stress ranges resulting from thicker pipe ends, as well as to reduced Stress Concentration Factors (SCF) resulting from forging process.

Consideration of vortex-induced vibration (VIV) and riser interference for a deep water riser array presents a complex engineering problem which requires knowledge not only of the behaviour of single risers in open-water current, but also shielding and VIV effects unique to riser arrays. The Genesis SPAR, located in 2,600ft water depth in Green Canyon 205, accommodates a closely spaced array of drilling, production and export risers run from a central moonpool. This paper describes riser VIV model testing and outlines how riser interference was assessed using a combination of wake flow modelling, global interference analysis and VIV prediction, calibrated to model test results and CFD analysis. Tow-tank model testing of riser pairs, performed at MIT, evaluated the influence of the wake field of an upstream riser on both VIV and drag for both bare and straked risers under different conditions of riser separation. A state-of-the-art approach to interference analysis was subsequently adopted which evaluated riser global response to loop current in the presence of both VIV drag-amplification on an upstream riser and wake-induced drag reduction on a downstream riser.Model tests showed that the use of helical strakes as VIV mitigation devices proved effective on single risers exposed to currents involving wide reduced velocity and Reynolds number regimes. Interference & blockage however, degrade the strakes' VIV suppression performance, as illustrated by results of RANS CFD simulations and tow-tank experiments of riser pairs. 1 CFD simulations performed by Professor John KallinderisInterference analysis of several loop current conditions illustrated that minimum riser separation was not associated with the highest surface current velocity, but with a lesser surface velocity which attenuated less with depth.Model testing clearly illustrates the impact of riser separation on the effectiveness of strakes.Interference analysis calibrated to such tests provides an innovative approach from which recommendations may be made about the conditions under which adjacent risers should be operated and mitigation measures available to avoid interference.

An examination of the corrosion-fatigue behavior of production quality welds in X65-type pipes was performed. Due to the low cycle operational nature of the production flowline system, the fatigue test frequency was substantially lower (0.01Hz vs. 0.33Hz) than typically utilized during corrosion-fatigue testing. Also the tests were performed at higher stress ranges than previous sour service fatigue tests, which to date have targeted riser fatigue loading regimes. Stress-life (S-N) samples were removed from segments of pipe with outside diameters of 10.75 inch (wall thickness of 1.30 inch) and 9.625 inch (wall thickness of 1.26 inch) containing fully inspected, production-quality circumferential welds. Environments examined included laboratory air conditions as well as deoxygenated brine supplemented by a gas mix of H2S and N2. For all environmental tests performed, the dissolved oxygen levels were maintained at less than 10 ppb during all testing. The measured fatigue life decrease in the curved pipe segments was in the range of 8–110 times due to the combined effect of the material and fluid property variables examined. The results of this work clearly illustrated the impact of sour-service corrosion fatigue, in welded carbon steel pipes, to the multitude of variables involved. Nevertheless, the foregoing experimental work clearly demonstrated the importance of performing environmental relevant testing when considering material and process selection for offshore applications.

As exploration and production move to even deeper water and more severe environment, the need to have a methodology for analyzing risers for in-line VIV fatigue damage without undue conservatism increases. The methodology presented in this paper reduces the conservatism in available methods by accounting for (1) the power-in region, (2) the power-out region (hydrodynamic damping), (3) competing modal excitation in the case of multiple mode excitation, and (4) the multiple constraints, if available, in the riser that result in irregular modal shapes. This methodology requires the use of a cross-flow VIV code with sheared flow capability such as SHEAR7, VIVA, or VIVANA. In this methodology the riser over the current profile is split into sections of cross-flow excitation and sections that have potential for in-line VIV excitation only. The cross-flow VIV code defines the sections for cross-flow excitation. All sections are analyzed for in-line VIV with the cross-flow VIV code using the appropriate in-line VIV force coefficients and Strouhal numbers. The assumptions implicit in the cross-flow VIV code regarding power-in, power-out, etc., are assumed valid for the in-line VIV analysis. The in-line VIV coefficients used in the analysis reported in this paper have been obtained from laboratory data, and are functions of both the VIV response amplitude and reduced velocity. The coefficients have been modified to give in-line VIV response amplitudes with the methodology presented that are consistent with DNV-RP-F105. The fatigue damage along the riser represents the sum of the damages produced by in-line VIV excitation for each of the riser sections.