Abstract
At deepwater sites, skirted mudmats with a sliding mechanism have been the preferred choice to support pipeline or flowline end terminations (PLET or FLET) on soft clay seafloor to accommodate line expansions and contractions that occur during shutdown and restart cycles. The sliding mechanisms can experience line expansions of up to 4 m, or if it were to lock-up, horizontal forces in excess of 2,000 kN. The design and fabrication of sliding mechanisms for such large movements can often become challenging and costly. As an alternative, the industry is now designing mudmats that do not utilize sliding mechanism but rather slide directly on the seafloor, if necessary. While it is an attractive alternative and is gaining popularity, all aspects of direct sliding mudmat design are not sufficiently understood because of the complexity that arises from the interactions among the various components, i.e., flowline/pipeline, mudmat, jumpers, and the supporting soft seafloor behaving as one system. The design must ensure that the mudmat will not unduly sink, rotate or dig into the seafloor in the course of the back and forth sliding caused by many shutdown and startup cycles during the field's operating life. Therefore, it is essential to sufficiently understand the characteristics of the sliding motions and estimate as accurately as possible the resulting interaction forces in the flowline-mudmat-jumper-seabed system. This will allow the designer to avoid the risk of any system instability from incoherent component motions, or potential overstressing of connections.
This paper investigates the direct sliding behavior of the mudmat motions using realistic simulations of the flowline-mudmat-jumper-seabed system. The simulations rely on coupled Eulerian (soft seabed) and Lagrangian (flowline, mudmat and jumpers) finite element analysis (FEA) method. The simulations yield insight into the complex mudmat sliding behavior interacting with the jumper, the hydrocarbon line and the supporting seafloor. The simulations show that merely modeling the mudmat alone fails to capture the actual behavior of a real-life PLET/FLET's back and forth cycling on the seafloor, because for example, the restraints (and attendant stresses) at the flowline and jumper connections are not accounted for, or that slanting the sides of the mudmat reduces its tendency to self-embed decreasing connection stresses. The paper demonstrates that realistic simulations allow important visualization of mudmat interactions with flowline, jumper, and the seabed during the shutdown-startup cycles, and provide accurate determination of the associated critical responses.
