The Utgard template is a multi-well manifold template supported on four 7m diameter suction caissons that penetrate about 5m into layered soil with sand and clay. The observations during installation of the template provide valuable experience that have been used to calibrate and improve the methods to predict skirt penetration resistance in layered soil. The interpretation of the penetration data shows that the skirt penetration resistance can be calculated by the semi empirical bearing capacity approach proposed by Andersen et al (2008), using best estimate soil parameters based on CPT data and empirical correlations. The penetration resistance is, however, overestimated in some of the dense sand layers. This is believed to be due to reduced resistance caused by pore pressure gradients from flow of water, even if there are clay layers above and below these sand layers. The lower resistance is not seen in all the dense sand layers, and experience from other installations also indicates that such reduction will not always be the case in layered soils. It is thus uncertain when the reduction in penetration resistance due to flow of water can be relied upon in design when there are clay layers above the sand. The observations indicate that additional skirt penetration occurs when the underpressure is reduced to zero and increased again. This implies that the skirts can be penetrated by a lower maximum underpressure if the underpressure is cycled. The back-calculated parameters in the clay layers correspond to empirical factors in the CPT method that are higher than the DNVGL (2017) most probable recommendation for stiff North Sea clays.
Knut H. Andersen, Knut Schrøder, Viggo Karlsen and Egil Solhjell
Keywords: Skirt penetration, layered soil, measurement, interpretation, recommendation
Publisher: DFI