A reliable procedure for load coefficient determination in structural reliability assessment of ageing offshore platforms

Abstract

Most of the oil and gas jacket platforms in Malaysia have exceeded their design life span with various underwater structural irregularities. Through the bow-tie risk assessment approach, it is predicted that there could be potential threats of hazard due to the unreliable procedure used to determine the load coefficient (??) value which could contribute to the failure in structure caused by extreme wave-in-deck. This issue is attributed to the unreliable procedure used to determine a load coefficient (??) value for wave height maximum limit at value in limit state equation. In practice, a range of 1.7 to 2.0 of load coefficient (??) value of 1.7 is recommended for practical application without considering an alternative reliable procedure to determine the appropriate values for a specific location and type of structure. In addition, the current practice to determine the appropriate load coefficient (??) value is by site measurement monitoring which is very costly and inefficient for offshore works. The study herein aims to develop a new alternative reliable procedure for load coefficient (??) determination, particularly for structural reliability assessment of ageing offshore oil and gas jacket platforms. A risk-based assessment (RBA) has been widely practised by the industry and it is based on the design code for fixed offshore structures that utilize the probabilistic model approach on load model (wave load) and load strength (load resistance) of limit state equation. Global Ultimate Strength Assessment (GUSA), which has been developed by PETRONAS, is one of the methods used in the study to compare the probability of failure (??????) and return period (RP) against ISO 19902. The results demonstrate that the most reliable procedure of load coefficient (??) range from 1.7 to 2.1 with eight (8) percent in coefficient of variance (COV) for the load model method. The accurate load coefficient (??) value was determined by the structure’s experiencing wave loading by at least two (2) prescribed return period (RP) at the long-term probability distribution. The ratio between the proposed and standard practice of load coefficient (??) was determined and evaluated for the platforms studied. In this study, a comparison between standard practice and the proposed reliable procedures indicates that the standard procedure systematically overestimate the structural probability of failure (??????) by up to 74 percent. Meanwhile, the return period (RP) is significantly underestimated by the standard practice at five (5) times lower than the proposed procedure. Results also indicate that the structure configuration, subsidence effect and extreme water level influence the selection of load coefficient (??) value. The results generated comply with the standard compliance of value delivery and classification of benefits to the platform operator and thus, are beneficial economically in terms of resources optimisation and platform’s reassessment.