Narrow-banded, well-organized vortex shedding associated with relatively low frequency vibration and relatively high amplitudes at lower wind speeds (8-18 mph) has resulted in several pipeline weld failures due to high cycle fatigue attributed to primary mode windinduced vibration (WIV) in the Alaskan Arctic. It has been postulated that the reason the secondary modes (2 lobes per span) are not excited by vortex shedding is that they require a span-wise correlation of lift forces that is much more difficult to achieve than for the lower frequency primary modes (with 1 lobe per span). The higher frequency modes are also associated with higher wind speeds which occur less frequently and higher Reynolds numbers. However, for pipeline configurations with very low vibration frequencies and/or small (aerodynamic) diameters, it is possible that secondary modes can be excited because they are associated with wind speeds in the subcritical Reynolds number regime. Alaskan Arctic North Slope oil and gas operators have employed successful mitigating measures (pipeline vibration dampers and tuned vibration absorbers) to prevent potential severe fatigue damage to pipeline welds caused by these lower wind speed events and the associated primary mode(s) of WIV. However, no protective measures have been routinely employed to prevent pipeline girth weld fatigue failures from secondary mode WIV based on the low likelihood of occurrence.
In late 2004, a high cycle fatigue pipeline girth weld failure attributed to high frequency, low amplitude secondary mode WIV associated with higher wind speeds (30-35 mph) occurred. Analyses revealed high quality welds, with no significant weld defects or metallurgical flaws that would have caused this failure had it not been for the cyclical stresses caused by the WIV. Significant secondary mode WIV was not predicted to occur by the analytical models historically used to analyze WIV of North Slope pipelines.
