ABSTRACT

Estimates of rare wave heights and crest heights in the Gulf of Mexico were derived from the Ocean Data Gathering Program hindcasts of twenty-two severe hurricanes affecting the Gulf since 1900. Results vary little between the sectors of the Gulf of Mexico investigated, but there is some fluctuation on a site by- site basis relative to composite sector values. Rare wave heights are not sharply focused by direction. Wave height results are not strongly influenced by the statistical method used, but crest height results are. Results in this paper provide a basis for evaluation of platform design wave criteria in the Gulf of Mexico.

INTRODUCTION

The parametric representation of rare storm waves is an important basis for static wave force calculations and for the selection or minimum deck clearances in the design of fixed offshore structures [1-3]. The purpose of this paper is to present-estimates of rare wave heights and crest heights in the Gulf of Mexico based on the Ocean Data Gathering Program hindcasts of twenty-two severe hurricanes affecting the Gulf since 1900. Gulf of Mexico operators can compare these results with their present platform design criteria.

In order to accurately assess return periods for large, rare waves produced by storms, three things are required:

  1. an accurate storm sea state generation model;

  2. knowledge of the conditional probabilities of individual wave parameters in a random sea; and

  3. statistical analyses.

Although the specific results are applicable only to the Gulf of Mexico, the philosophy and procedures employed are applicable to any area where historically severe storms are the predominant environmental hazard associated with fixed platform design.

Figure 1 illustrates conceptually what is done in the historical analysis of severe storms. The universe of all possible storms that affect a platform location contains those storms that have occurred (past exposure) and those storms that will occur (future exposure). An underlying assumption of the analysis is that the statistics of events that have occurred will be representative of the statistics of events that will occur. The subset of past storms that are judged to have been the most severe based on meteorological parameters forms the basis for the hindcast study. Wave data obtained for a small subset of past storms provide a basis for hindcast model validation. Hindcast technology for severe hurricanes occurring in the Gulf of Mexico is available through the industry sponsored Ocean Data Gathering Program [4]. Tropical storm sea state data is available for the period between 1961 and 1974. The ODGP hindcast model adequately predicts the measured sea states given the conventionally available meteorological parameters of the storms [5].

Previously published results are available which have been based on different hindcast models and different statistical procedures [6,7]. This paper and a paper by Ward, Borgman and Cardone [8] present results obtained with the ODGP model.

The information developed in this study is presented in four main sections: discussion of data base, statistical treatment of hindcasts, wave parameter results and sensitivity studies.

This content is only available via PDF.
You can access this article if you purchase or spend a download.