Stephen Rose

Journal Paper

A Probabilistic Structural Reliability Assessment of Existing North Sea Platforms

ASME International Conference on Ocean, Offshore and Arctic Engineering

There are four key areas in the SRA of fixed jackets that have been impacted by recent advances in our understanding of extreme offshore wave conditions. These are: the probability distribution of crest heights in extreme seas; the occurrence of breaking waves and their associated enhanced near-surface kinematics; the calculation of wave-in-deck loading where inundation occurs; and, the probabilistic incorporation of multiple structural failure modes. The authors have applied the new approaches for platforms in the Central and Northern North Sea to assess the impact on the estimated structural reliability. The paper describes the experience of using the new methods on one of these platforms in terms of: the sensitivities of the approach to various assumptions; the simplifying assumptions that were used to make the problem more tractable; the computational complexity associated with adopting the full probabilistic approach that is inherent within the new methods; and, the overall impact of using the methods on the calculated structural reliability.

Graham Feld, David Randell, John Kent, Stephen Rose, Emmanuel Fakas

Journal Paper

Examining Field Measurements of Deep-Water Crest Statistics

ASME International Conference on Ocean, Offshore and Arctic Engineering

This paper concerns the field measurements of wave crest elevations and wind speed obtained from fixed platforms in the North Sea during the period of 2011 to 2015. An improved quality control process is proposed by the authors aiming to maximise the amount of waves that are not discarded, whilst maintaining a reliable database. Applying this revised methodology, the total number of waves in the quality-controlled dataset is approximately 184 million. In total, the measurements correspond to 32 years of continuous data and contain approximately 0.0054% of rogue waves, according to the definition of Haver (2000). As the sea-state steepness increases, it is observed that the measured crest elevations are poorly modelled by the Forristall (2000) distribution; similar behaviour is observed for sea-states with high wind speeds. The model by Tayfun and Fedele (2007) is shown to be a better alternative in these cases, demonstrating that non-linear amplification beyond second-order cannot be neglected.

Isavella Makri, Stephen Rose, Marios Christou, Richard Gibson, Graham Feld

Thesis

Extreme Waves in Wind Driven Seas

Imperial College London

Estimating the probability associated with the main properties of large waves has key engineering significance for the design of vessels and offshore structures. This paper describes the measurement, quality control, and analysis of field data from 17 installations across the North Sea and South China Sea. Combined, the processed dataset contains 929,740 sea states and 202 million waves. Examination is made of the crest elevation and wave height statistics, finding good agreement with the theories of Forristall and Boccotti, respectively. Further, the accuracy of common theoretical crest elevation models are presented as functions of various sea state parameters. The data contains 10,799 rogue waves and a 5-day storm event is presented, containing several of the largest normalised crest elevations. The presence of high, sustained steepness and wind speed appears linked to a moderate increase in the frequency of rogue waves. Variations in wave shape are illustrated as functions of wind and wave parameters. Both the crest period and crest sharpness are seen to increase with significant wave height and, to a lesser extent, the wind speed. The spectral peak period shows a clear link to the upper-bound of individual crest periods.

Stephen Rose — supervisors: Marios Christou, Richard Gibson