Risers, pipelines and flowlines for deep water applications are subject to corrosive environments. Especially, in the presence of hydrogen sulfide which makes the field sour, their fatigue performance becomes significantly degraded. In order to quantify the sour degradation effect, a knock-down factor has been introduced. This factor is defined as the fatigue life reduction relative to the in-air fatigue life.
Several sets of fatigue test results in sour service environments have been published. These include strip specimens of different sizes, e.g., diameters, wall thicknesses, and arc lengths. Naturally, the knock-down factor must be based upon a statistically valid number of fatigue test results obtained from the same specimen geometry and the same loading conditions tested in air and in sour conditions. Currently, the database available in the open literature is too limited to properly define a knock-down factor. Moreover, there is a great deal of scatter within the database and each test in a sour environment is costly and time consuming. Thus, it is difficult to establish a statistically valid database upon which to base the knock-down factor.
A mesh-insensitive structural stress method has been developed by Battelle researchers and has been proven to be highly effective in correlating the fatigue behavior of welded joints. In 2007, the Battelle structural stress based weld fatigue master S-N curve was included in ASME Section VIII Div. 2 because it successfully consolidated more than 800 fatigue test results for weld toe failures onto a single master S-N curve with very little scatter, regardless of specimen shape, size, loading type, and steel alloy [1–2].
A knock-down factor is derived by applying the Battelle structural stress method to the existing database for sour environment tests and by using the current in-air database as the reference condition. This approach will reduce the uncertainty in the knock-down factor because it allows a wider range of sour environment data from specimens of different sizes, types, and loading conditions to be combined, while simultaneously reducing scatter. As such, a unified knock-down factor can be determined with greater statistical validity and wider applicability for design recommendations in sour conditions.