Analysis of Extreme TLP Tether Tension Amplitudes With Comparison to Model Tests

The Snorre TLP has been operating in the Northern North Sea since 1992. During the last year, an extensive reanalysis of tether tension, TLP motion and airgap has been carried out in order to ensure that the TLP can withstand updated environmental conditions and design requirements. It is the object of the present paper to describe the analysis methods which have been employed to determine extreme tether loading and to compare analysis results with model test results. The model tests were carried out in the facilities at MARIN in the Netherlands and considered more than 200 three hour realizations of irregular seastates with and without wind and current. Typical fatigue seastates as well as extreme 100 year and 10 0000 year seastates were investigated for both head and quartering seas. The extensive database of model test results gives an opportunity to study the accuracy of the analysis methods. The analysis of the TLP response was carried out in the time domain using the SIMO software developed by Marintek. In the work described here, the results from a purely linear radiation-diffraction analysis is used as the main load model in SIMO together with slender element additions based on the incident wave kinematics. The slender elements give loads on tethers and risers, viscous loads on the TLP hull and also provide a simple third order load model for the excitation of high frequent response in the surface zone. SIMO has also been implemented with second order quadratic force transfer functions, but it is concluded that the second order loading does not give significant contributions to extreme tether tension. It is concluded that, at least for the Snorre TLP, the relatively simple analysis model gives good agreement with measured extreme tether tension provided that the TLP is modeled carefully. By filtering the tether tension time series, the contributions due to wave frequent and high frequent (resonant) tether tension may be separated. It is shown that the main discrepancy between model test and analysis results of tether tension lies not in the magnitude but in the simultaneity of wave frequent and high frequent tether response. If this is due to the tendency for large waves to propagate towards the front of the nonlinear wave envelope, this may be extremely difficult to capture in analysis models.