Abstract A197

The objectives of this API sponsored research project were:
(1) to analyze the variation and sensitivity of tie-down loads for drilling rigs and their substructures on different deepwater floating drilling and production systems, and (2) develop information that can be used to develop guidance and recommended practices for designing tie-down systems for drilling rigs on various types of floating drilling and production systems.

During hurricanes Ivan, Katrina, and Ike, several drilling rigs on floating production systems (FPSs) and Tension Leg Platforms (TLPs) were moved, and in some instances, even toppled.  An initial study of the failures during hurricane Ivan was funded by the Minerals Managements Service (1-4).  Computer programs were developed and used to estimate the forces on bolted clamp tie-down systems during hurricane Ivan.  The failure modes of bolted clamps were studied and modeled.  The hurricane loads were compared to the failure capacities of the clamps in slip, bolt tension, and bolt shear.  Results indicated the sensitivity of clamp loads and failures to structure accelerations as well as wind loads, and thus the importance of purpose-designed tie-down systems for the specific structure-drill rig combination and function (derrick/drill floor tie-down or drilling substructure tie-down systems).  Slip was identified as the most likely failure mode, which was not inconsistent with observations during Ivan.

The maximum loads on the tie-down footing were simulated in hurricane environments. In this API funded study, the MMS study was expanded and extended.  The focus was on the loads on the tie-down footings and not the loads for a specific tie-down system such as the bolted clamps studied in the MMS project.  The more general approach allows the results to be used in developing guidelines for designing all types of tie-down systems (e.g., bolted clamps, other types of mechanical or hydraulic clamps, weldments, mechanical stops or pins, etc.).

Loads on the following structures-drill rig combinations were simulated:

• TLPAA – a TLP in 3000 ft with a drilling rig AA (representative derrick & substructure for a TLP)
• SparAA – a Spar in 3000 ft with drilling rig AA
• SemiAA – Semi in 10,000 ft with drilling rig AA
• SparAS – Spar in 3,000 ft with drilling rig AS (representative derrick & substructure for a Spar)

Each structure and drilling rig combination was analyzed for hurricane wind, wave, and current conditions that represented 100-year, 200-year, and 1000-year return periods as specified in API 2INT-MET (5) for the Central region.  The time varying wind loads for a 3-hour period were simulated based on the API wind spectra.  The time varying global accelerations for the floating structures were simulated for a 3-hour period using the TAMU-WINPOST model, which has been verified through numerous comparative studies against model tests and field measurements.

We had planned to simulate wind loads on the derricks and substructures using the improved techniques recently benchmarked by the API Spec 4F and 2TD Task Groups and now included in the new API Spec 4F (6).  However we were unable to obtain sufficiently detailed information on actual rig designs, so we resorted to using representative drilling rigs and simulated the wind loads from the available data.

The random time series of the loads on the tie-down footings were computed from the simulated wind loads and structural accelerations using the coupled structure and derrick model developed in the MMS study.

Forces on tie-down footings were analyzed to examine the differences due to the various structure-drilling rig combinations.   A simplified and unified relationship was established between the maximum simulated loads and the sum of the maxima of the wind, inertia, and gravity loads.  This relationship fits the results for all combinations of floating structures and drilling rigs studied.  That relationship was then used to develop a simple method to estimate tie-down footing loads for the 100-year design case and the 1000-year robustness check case.  The relationship seems to be sufficiently robust and tractable to be useful in providing design guidance for recommended practices.