The paper presents a study of the extreme response statistics of a tension leg platform (TLP) subjected to random seas. Two different approaches are compared: A numerical integration method based on saddle point integration and the Monte Carlo method. While the saddle point method is a mathematically attractive technique, which gives numerically very accurate results at low computational costs at any response level, the advantage of the Monte Carlo method is its simplicity and versatility. It is demonstrated in this paper that the commonly assumed obstacle against using the Monte Carlo method for estimating extreme responses, i.e. excessive CPU time, can be circumvented, bringing the computation time down to affordable levels. The agreement between the two approaches is shown to be remarkably good.

This content is only available via PDF.
You do not currently have access to this content.