Tension-leg-platforms (TLPs) are increasingly being used in offshore regions of high seismicity. The TLP foundation typically consists of very long driven piles supporting the TLP tendon loads. These piles experience significant uplift loads generated by the inherent system buoyancy. API RP 2T is considered the industry standard for TLP design which provides guidance on foundation design; however; available guidance on seismic design of TLP foundations is limited. The focus of this paper is to present and compare two types of seismic design methods for TLP piles and discuss associated outcomes, which should provide greater insight into the complexities involved in seismic design analysis.
The seismic analysis of TLP piles can be performed in two ways; coupled and decoupled analysis. In a conventional decoupled analysis, as practiced in the industry, the foundation pile is analyzed independent of the TLP structure considering various design loading conditions, including seismic loads. In the coupled analysis, the entire system including TLP structure, tendons, and piles are modeled and analyzed as one integrated system. This is done to investigate both system and component responses due to the impact of seismic ground motions. In the coupled approach, the system can be analyzed by the response spectra method for an extreme level earthquake (ELE) and the dynamic non-linear analysis for an abnormal level earthquake (ALE). The effects of seismic kinematics and inertia, soil degradation due to cyclic loading, soil damping, etc. are considered in both types of analyses.
Two case studies are presented where these methods have been applied for two TLP projects located in seismically active regions. The outcome of the study is to present a comparison between the conventional decoupled approach and the proposed coupled seismic analyses of TLP piles. This comprehensive seismic study aims to provide detailed insight into the seismic design methods of a TLP foundation to supplement the design guidance presented in API RP 2T. Although complex and time consuming, the proposed coupled approach provides greater insight into the system response.