No abstract

The height of waves at North Sea oil and gas installations is an important factor governing the degree to which operational activities may be undertaken at those facilities. A link between the North Atlantic Oscillation (NAO) and winter (defined as December-February) wave heights at North Sea oil and gas installations has been established. A tool has been developed that uses a forecast NAO index to predict the proportions of wave heights in four categories that could be used to assess the operational downtime that will be experienced in the coming winter. The wave height forecasting system is shown to have useful skill in predicting the probability of occurrence of a stormy winter, and therefore probability forecasts provide a potentially useful guide to whether more or less disruption than the ''climatological mean'' might be experienced. The main limit on the skill of the wave forecasts is our very limited ability to accurately predict the NAO index on seasonal time scales.

The accurate prediction of extreme wave heights and crests is important to the design of offshore structures. For example, knowledge of the extreme crest elevation is required to set the deck elevation of the topside of a jacket structure. However, methods of extreme value analysis have an inherent bias, and the manner in which they are applied affects this bias. Furthermore, there is uncertainty in the design parameters at the time of design and the possibility that the predictions will change during the life of the structure. This paper is concerned with the accurate prediction of design values that incorporate uncertainty. In the first part of this paper the details of commonly applied extreme value analysis techniques are examined. This is achieved through analysis of simulated data of known distribution. In particular it is the application of least squares minimisation routines that is investigated; however, comparisons are made with maximum likelihood estimation. From this, preferred approaches to the analysis are recommended and their advantages and disadvantages discussed. The methods are applied to the analysis of a North Sea data set and the implications for the design values ascertained. In the second part of the paper Bayesian inference is used to consider the effect of uncertainty in the predicted wave heights and crest elevations. The practical implications are determined by the analysis of a measured North Sea data set.

No abstract

The Faroe Islands are planning to open their territorial waters to oil exploration and production in 1998, when they have their first licensing round. The Faroese economy is presently dependent upon fisheries, with fishing, fish farming and processing accounting for almost 95% of all exports. The seas surrounding the Faroes are renowned for their cleanliness, and are of importance for cetaceans and seabirds as well as fish. In a unique collaboration, 23 oil companies have already started to work together in conjunction with the Faroese Petroleum Administration and Institutes in order to gather data on the environment around the islands. The main objective of the joint industry/institute group is to gather sufficient data to enable operators to carry out environmental risk assessments for future exploration drilling offshore the Faroe Islands. This paper describes the work and philosophy of the Faroes GEM, whose workgroups are studying the Geotechnical, Environmental and Metocean aspects of prospective areas of interest on the Faroese Continental Shelf.