Insights in the Application of Structural Reliability Analysis (SRA) for Challenging Pipeline Lateral Buckling Design

Author(s):  
Elena Gerginov ◽  
Andrew Rathbone ◽  
Chris Sullivan ◽  
Terry Griffiths

Lateral buckling of pipelines is influenced by, and sensitive to, a complex combination of factors associated with considerable uncertainty. Key elements such as pipe to soil friction, vertical and horizontal imperfections can vary significantly not only along sections of the same pipeline, but at the same location over time. These factors carry a probabilistic likelihood of occurrence that results in a probabilistic distribution of buckle formation forces and locations, as well as post buckle loads, as described by Rathbone et. al. (2008). To effectively capture all possible combinations of inputs and results, and thus gain a full understanding of potential buckle behaviour and interaction, would require many individual finite element analyses to be undertaken, which is not viable. The paper describes new insights into challenging pipeline lateral buckling design gained through use of a highly adaptable Structural Reliability Analysis (SRA) method. It discusses the difficulties in defining pipeline and seabed complexities while still retaining adequate detail to fully describe the pipeline system. A range of realistic lateral buckling design scenarios will be discussed, illustrating how to reduce existing limitations in probabilistic design through the use of the SRA method. The method is flexible enough to enable varying levels of detail to be modelled, from simple idealised checks of buckling susceptibility, which are commonly performed during the early phases of design, to the more complex analysis of pipeline failure, performed during detailed design. The paper will demonstrate the ability for the method to capture complex real life scenarios such as the effect of a bi-modal distribution in observed pipeline embedment, as well as the influence of covariance between frictions, on the probability of buckling and resulting strains.

Author(s):  
Marcus McCallum ◽  
Andrew Francis ◽  
Tim Illson ◽  
Mark McQueen ◽  
Mike Scott ◽  
...  

Approximately 1450 km (900 miles) of a 4020-km (2500 mile) natural gas pipeline system operated by Crosstex Energy Service L.P in Texas are subject to the Texas Railroad Commission’s (TRRC) integrity management rules. Consequently, in preparation for the construction of an extensive and robust integrity management program, Crosstex commissioned Advantica to assist in the development and application of a pilot study on a 13.4 km (8.3 mile) section of a 14” pipeline. The purpose of the study, which is based on Structural Reliability Analysis (SRA), was to compare the level of integrity that could be inferred from the use of Direct Assessment (DA) techniques with the level that could be inferred from ILI results. Based on a preliminary assessment of available data, the study identified both external and internal corrosion as potential threats to integrity. SRA was used in conjunction with ‘Bayesian Updating’ to determine the probability of pipe failure due to external corrosion, taking account of results from above-ground measurements and a number of bell-hole excavations. The above-ground survey techniques utilized included Close Interval Survey (CIS) and Direct Current Voltage Gradient (DCVG). A similar approach was adopted to address the threat due to internal corrosion, but hydraulic modelling was substituted for the above-ground measurements. A third study based on SRA was used to determine the combined probability of failure due to both internal and external corrosion taking account of ILI results. The outcome of the analyses demonstrated that the level of integrity that could be inferred from the use of Crosstex’ DA methodology was similar to that which could be inferred from the use of ILI. The results were presented to the TRRC for review and approval. This paper gives a detailed description of the SRA based methodology that was employed by Crosstex and presents the results that clearly demonstrate the comparability of ILI and DA for the purpose of integrity management.


Sign in / Sign up

Export Citation Format

Share Document