The Reliability Of Offshore Structures And Its Dependence On Design Code And Environment

Abstract

ABSTRACT The probability of failure due to extreme environmental overload is calculated for Tern, a manned self contained production drilling platform located in the northern North sea. This demonstrates its high reliability. Comparison with results obtained for structures in other areas designed to the same API-WSD code, shows that consistent levels of structural safety are not ensured. In contrast, this can be achieved using API-LRFD together with location specific load factors and joint-metocean design conditions, To assess the consequences of designing to API-LRFD, the structure is re-sized using this code and a design environmental load corresponding to response-based or joint-metocean design conditions as recommended in APl-LRFD. The reliability of the re-sized substructure is evaluated and the weight saving is quantified, We conclude that a rational application of LRFD results in high reliability and should allow location dependent factors on environmental load. The means for estimating these factors is presented. INTRODUCTION In this paper we study the relationship between the reliability of offshore structures and their environment and design code. The reliability analysis presented follows well established principles [1,2], but uses significantly more realistic models. It involves use of a hindcast data base and advanced oceanographic, wave loading and pushover analyses in order to quantify the probability of structural collapse due to severe weather. The models have been tested by predicting actual survival and damage of structures in the Gulf of Mexico that were exposed to severe hurricanes [3,4]. It has provided a rational basis for decisions concerning the evacuation of personnel from a Iiving quarters platform in the southern North Sea [5] and the use of jack-ups in the central North Sea [6]. Here, the calculation of the probability of failure of the Tern structure, under environmental loads is presented. It is an important milestone in reliability analysis. The Tern platform is instrumented. Records of winds, waves, currents and global loads during extremely severe environmental conditions have provided a full-scale validation of the "new wave" load models [7l. This not only provides a particularly sound foundation for the present work and previous studies [3,4,5,6], but also supports the significant reduction of environmental design loads for North Sea structures. In the API-LRFD design code [8] the use of extreme response statistics is recommended for obtaining 100 year environmental criteria. To investigate the consequences of this for the design of structures in the North Sea the existing Tern structure was redesigned for response based environmental criteria which lead to a 100 year load significantly less than the original design load. The primary objective of the present paper is to identify whether the API-LRFD design code, including the use of joint metocean design conditions, can insure reliability levels that are sufficiently high and consistent for different environments in the North Sea. A second objective is to identify the potential for weight saving.