Abstract

In recent years, spars have become a preferred industry solution for certain offshore deepwater developments. Perdido is the first spar platform to be operated by Shell. The Perdido spar has been recently installed in the ultradeepwater Gulf of Mexico Alaminos Canyon and is scheduled for first oil in 2010. This Direct Vertical Access (DVA) spar will operate at a water depth of 7,825′ and will be the deepest spar production and drilling facility in the world. Numerical predictions of the spar global motions in waves, wind and current are presented in this paper. Motivation for this study comes from two facts: 1. Each spar platform design is unique in terms of its size, number and geometry of heave plates, riser system and mooring system. 2. Metocean design criteria have been increased in view of the recent hurricanes. Model tests of the Perdido spar were conducted at MARIN at a scale of 1:59.94. In these experiments, several Gulf of Mexico (GOM) wave, current and wind environments were considered. The six-degree-of-freedom motions, deck accelerations, air gap, as well as the loads on the heave plates, mooring lines and risers were measured. In this paper, global motion predictions of the Perdido spar are given using Shell’s in-house COSMOS/ WAMIT suite of programs. Extensive comparisons between the numerical predictions and the experimental results were undertaken. In all cases, the comparisons are very good. In order to include heave viscous loads and damping, special line members were included at the bottom of the hard tank, the bottom of the soft tank and each heave plate, in addition to standard line members used to describe the truss. These special members contribute heave viscous loads with drag coefficients selected from the Perdido experiments. Several heave plate configurations were considered to systematically study the impact of heave plates on the spar global motions. The influence of the heave plate geometry on the heave added mass and on the global motions was derived using WAMIT. The strakes’ actual geometry was also included in the WAMIT diffraction analysis. Most of the moonpool area at the bottom of the Perdido hard tank is closed. As a result, the pumping mode was not excited during the experiments. However, numerical simulations with WAMIT showed a sharp peak at the “pumping mode” resonant frequency. This peak was suppressed by introducing a second floating body that capped the moonpool at the water surface. Based on these learnings, recommendations for global motion modeling are presented in this paper.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call