Numerical assessment of preloading strategies for jack-up vessels in cohesive soils

Abstract

As the demand for renewable energy is constantly rising, more offshore wind farms are being built at sea to meet European targets towards sustainable energy transition. Van Oord’s offshore installation vessel Aeolus was purposely-built to transport and install foundations of offshore wind turbines. To ensure that the vessel remains competitive in a market that is changing rapidly, Van Oord decided to upgrade this vessel to expand its operational capabilities to work on soft soil seabed conditions.

Equipped with moveable legs, the mobile jack-up vessel is able to elevate its hull out of the water and provide a stable work platform. Prior to each installation, the soil is preloaded by penetrating the spudcan footings into the soil under the self-weight of the structure, followed by progressively pumping sea water into the ballast tanks until the target preload is reached. Offshore industry guidelines follow the conventional bearing capacity theory to assess jack-up procedures. However, these conventional procedures are inadequate in incorporating all relevant mechanisms that contribute to the preloading behaviour in cohesive soils. The effect of consolidation around the spudcan footings on the soil strength and the viscous strain rate-dependent soil strength are two important aspects which are poorly described in the current guidelines. Using numerical modelling, a better understanding of these mechanisms is aimed for in order to evaluate spudcan bearing capacity in cohesive soils in a more detailed design.

An axisymmetric finite element model of the spudcan foundation is constructed in PLAXIS 2D to assess the effects of consolidation and rate-dependence of soil strength on the use of different preloading strategies. As spudcan penetration is relatively fast and clay permeability is relatively low, consolidation effects are found to be insignificant during the preloading process. Contrary, viscous strain rate effects dominate the soil strength of clay in predominantly undrained conditions. Higher shear strengths are found for higher spudcan penetration rates. During the standard jacking process, when the target preload value is reached and the spudcan is stopped, a reduction of the leg load occurs due to decreasing viscous resistances, which must actively be brought back to the target preload value. In clay soils, this leads to further penetrations and multiple preload cycles are required before the spudcan response is stabilised. In terms of preloading duration, a rather slow spudcan penetration rate is therefore beneficial when using the standard jacking procedure.

Alternatively, a different jacking procedure is assessed numerically, which adapts for these viscous strain rate effects by overshooting the target preload value by 10-15%, in order to reach the stable spudcan response more quickly. In this way, the spudcan is brought to the depth, which would have been reached by the creep-like additional penetrations in the iterative preload cycles under the target preload value. Once the load is reduced from the overshoot to the target preload value, it is likely that the preload can be sustained without additional penetrations and preload cycles. The overshooting jacking procedure reduces the required preloading time considerably for both spudcan penetration rates. The largest time savings are gained for the largest penetration rates. However, using this method, larger leg penetration predictions compared to the standard jacking procedure are observed for decreasing spudcan penetration rates. That is because in the standard jacking procedure, less creep-like additional penetrations are present when using a slow spudcan penetration rate. The overshooting jacking procedure can thus be considered beneficial for relatively fast spudcan penetration rates, both in term of preloading duration and spudcan penetrations.