A fundamental investigation of the effects of the hydrostatic pressure test on the structural integrity of pipelines using structural reliability analysis

摘要

For many years the hydrostatic pressure test has been performed as a final integrity check of pipelines. The test involves subjecting the pipeline to a hydrostatic pressure that is typically some specified percentage (125%-150%) of the maximum allowable design pressure or which results in a hoop stress that is a specified percentage (90%-105%) of the specified minimum yield stress. The test thus provides a margin of safety against failure under the operational pressure loading. In doing so the test not only demonstrates that the material and geometry are adequate but also provides some confirmation of the workmanship levels of the welded joints. However, over recent years the value of the hydrostatic test has been the subject of a continually broadening debate. It is now being argued that the information provided by the test can be obtained by other means. For instance, the quality of the material and geometry have improved due to improvements within the pipe mill and advances in inspection techniques can lead to higher standards of weld quality. Whilst the above and other improvements are widely accepted it has generally not been possible to quantify their effect on the structural integrity. This paper describes a study that focuses on this issue. The approach adopted here is to quantify the effect of the hydrostatic test on relevant failure modes using structural reliability analysis. The outcome of the structural reliability analysis is a probability of failure, which is determined from a systematic consideration of the elements of uncertainty associated with each failure mode. The effects of the test are assessed by a detailed comparison of the failure probabilities resulting from the application of the structural reliability analysis to a range of scenarios comprising the effects of workmanship, weld inspection and hydrostatic test pressure. The results allow a view to be formulated on conditions under which it may be possible to forego the test or, conversely, to identity other mitigating factors which may be relaxed in view of the test.