Ice Gouging and Its Effect on Pipelines

Abstract

Abstract The objective of this paper is to demonstrate how modelling can be used to account for the inherent correlations while assessing the pipe response to ice gouging – and thus narrow down the uncertainties associated with the pipeline design process. A custom-developed numerical model of ice gouging has been developed and exercised to better understand how the environmental conditions affect the gouging process. Further, numerical simulations of the keel-soil-pipe interaction have been performed, relating to the input and output of the ice gouging model. The ice gouging simulations quantitatively demonstrated the effect of the governing parameters on the gouge depth. Geotechnical conditions are very important as the main source of resistance against the driving force from ice, making a noticeable difference in the gouge depth. Force balance is also important, in particular how the vertical forces are generated or/and limited by natural phenomena such as the shape of an ice feature, seabed topography and tides. The effect of the driving force and the keel resistance limits have not been dealt with within the scope of the present study. The ice gouging simulations demonstrated that ridges with steeper rake angles result in deeper gouges. Similar gouge depths have been attained irrespective of the path – via water level changes or via seabed slope - as long as the driving force was available. The keel-soil-pipe simulations demonstrated that increasing the rake angle results in lower pipe response. Deeper gouges give larger effect at the same clearance between the top of the pipe and the gouge bottom. Considering the performed simulations jointly, it can be concluded that selecting the n-year gouge depth implicitly sets the conditions for assessing the variability in the additional governing parameters. This is mainly applicable to the keel's rake angle as the factor having the strongest correlation with the gouge depth, keeping other parameters unchanged. Appreciating the correlation above and its effect on the n-year pipeline load effect can prevent possibly overly-conservative parameter combinations. It is argued that steeper keels have the highest potential to produce deepest (extreme) gouges. For these conditions, the extreme gouge load effect seems to be acceptable as long as there is a gap between the pipe and the ice keel and the pipe wall thickness is relatively large. Thus, the pipeline design against ice gouging load effects is likely to be governed by the maximum gouge depth, as the gap itself is found to define a sharp limit between an acceptable and an unacceptable design condition. In turn, this sets high requirements to reliable determination of the gouge depth.