Numerical & experimental investigation on the effect of scour formation and protection on the monopiles behavior

Abstract

Offshore wind farms are rapidly developing in Western Europe as an alternative means of clean energy. The foundation of wind-turbines in the shallow water of the Northern Sea is almost exclusively (over 87%) performed by short rigid monopiles, as an efficient and relatively cheap solution. Scour formation, meaning the soil removal around the pile due to the actions of the waves and currents, is a potential hazard for the functionality and the integrity of the structures supporting the turbines. This thesis investigates the effect of scour formation in the soil-pile capacity and examines the possible contribution of the scour protection layers in the stiffness of the system both under static and cyclic loading. The methodology followed included 18 main static and cyclic centrifuge experiments and 9 basic cyclic numerical analyses. The physical modelling included an aluminum pile, equipped with strain gauges in order to calculate bending moments and ultimately derive the “p-y curves”, which was embedded in a dense dry Geba sand. Numerical analyses have been conducted in PLAXIS 3D, with a symmetrical fully drained model, in prototype scale, with the same set of materials (Geba sand, aluminum pile).
The results have shown that scour formation can diminish the lateral soil capacity in the ultimate limit state (ULS). Scour depth is the most critical characteristic of the scour hole geometry, but the scour width at a specific depth can also make the difference between structure’s integrity and failure. Therefore, the term “local” scour is deemed insufficient, as it cannot be described by a unique geometry. A valid design against scour is recommended to include a series of analyses with combinations of scour depth and width in realistic ranges. Experimental derived “p-y curves” have been compared with the ones proposed by the API method, concluding that the API overestimated both the initial response and the ultimate capacity of the soil reaction. It is proposed to update the existing regulations for the rigid piles and add rotational springs to simulate the effects of the considerable shear developing in the tip of rigid monopiles. The effect of the scour width in the “p-y curves” was observed to be limited in the shallower depth of the pile, as going deeper resulted almost to the same soil response regardless of the type of scour. The cyclic centrifuge experiments focused on the load type, investigating different scenarios of “one-way” and “two-way” load patterns. It was shown that the “one-way” case is more favorable in terms of accumulating deformations compared to the “two-way” case which experienced higher residual displacements, as long as the same maximum load was applied. This was attributed to the smaller dissipation of energy and hence destruction of the soil structure by the “one-way” loading. However, when the maximum load applied in a “two-way” experiment is considerably smaller than the equivalent of the “one-way” test, smaller deformations observed in the “two-way” test, implying the significance of the maximum load. The last section of this thesis, the numerical analyses, focused on the scour protection effect in the mechanical properties of the soil-pile system when subjected to cyclic loading. It was shown that a typical protection layer can highly increase the stiffness of the system and hence decrease the accumulated displacements. The length of the protection layer is not crucial, as change in its magnitude does not alter considerably the reduction of the accumulated displacements, in contrast with the thickness which has a larger impact on the stiffness of the system. It is concluded that scour protection layers can considerably increase the soil resistance around the monopile, allowing for smaller embedment length, so their contribution in the soil-pile stiffness should be taken into account for a more economic design.