Abstract

In this presentation, it will be described which approach was taken to specify global and local ice loads for the structures in two different oil & gas projects in sub-Artic areas: the Sakhalin II project in the Sea of Okhotsk and the Kashagan project in NE Caspian Sea. In both projects, the approach was a combination of deterministic methods described in the different design codes, probabilistic methods and scale tests in an ice tank.  For the Sakhalin II project, global ice loads were particularly relevant for the design of the two multi-legged platforms (Piltun-B and Lunskoye-A). In contrast, in the NE Caspian, an accurate determination of the global ice loads was mainly relevant for design of the ice protection barriers rather than for design of the islands itself. Instead, much more attention was paid to specifying the risk of ice encroachment onto the islands.  In all cases, it was found that the limited amount of ice data often resulted in a conservative design with still considerable uncertainties in the design loads.