The transition to more sustainable energy has led to a growing demand for offshore wind energy, necessitating larger structures and heavier foundation piles. During pile installation this increases the risks of uncontrolled pile run, which can have fatal outcomes and project delays. Offshore superstructures predominantly rely on largediameter open-ended driven piles. Intermediate soils pose significant challenges in determining soil resistance during driving and pile run, with calculations typically performed using Static Soil Resistance to Driving (SRD) methods. The Alm & Hamre method is preferred for deep foundation piles in mixed soils due to its inclusion of friction fatigue.
The goal of this research with corresponding research objective is to improve pile run velocity and trajectory predictions for offshore open-ended pile installation in intermediate soils. This objective is reached through research on drainage state, identifying soils susceptible to a shift in this drainage state, and analyzing velocitydependent resistance for CPT and pile velocities. The findings are then incorporated into a modified SRD model. The results are compared to a case study using CPT and borehole data as input, and installation video’s and driving data as validation material.
Key findings indicate that with increased pile velocity the drainage state of several soils can shift towards the more undrained spectrum and therefore the soil will have a smaller soil resistance. These soils with a lower soil resistance during pile installation velocities then predicted include intermediate soils such as silt, sandy silt, and silty clay. Thin alternating layers of sand, clay, and silt are also likely to experience a shift in drainage state. Later silty sand is identified as a soil with a high possibility of being prone to such drainage state shifts.
The SRD method, incorporating velocity-dependent resistance, predicts pile run 31% more accurate than models without this consideration. By including velocity-dependent resistance drops, the model accounts for the changes in soil resistance that occur during pile run, leading to more accurate predictions compared to the standard SRD model. The model used in this research uses a single SRD update for velocity dependent resistance. However, in scenarios with large pile runs trajectories and high pile velocities, or when a substantial portion of the soil is prone to a drainage shift, performing a single update for velocity-dependent resistance will not result in a converged solution. As such for a correct solution, multiple iterations are necessary.
When the model predicts a deeper Self weight penetration depth than observed, the predictions for pile run initiation are not reliable. Given that pile run initiation can be very delicate, further research is needed for locations with CPT and borehole data directly beneath the pile. Additionally, incorporating hammering parameters, such as the added weight due to hammer momentum, should be explored to improve these predictions.