Prediction of Pile Run During Pile Driving - Analytical Model and Field Observations

Abstract

Abstract In offshore construction, geotechnical risk assessment of pile installation during the preparatory phase is of major importance. Pile driveability analyses are part of these preparations and provide, primarily, insights into the relation between hammer blow count, driving stresses, and the static soil resistance during driving (SRD). In case the total weight of the "pile driving system" (i.e. the combination of pile, follower and hammer) exceeds the SRD, a phenomenon referred to as "pile run" or "drop fall" might occur. Usually, at this point in wave equation analysis software, no hammer blow count is calculated until a depth is found where the SRD increases sufficiently to support the weight of the pile driving system again. For many (offshore) projects, with ever increasing weight of piles and hammers, the consequences of pile run to equipment and foundations could be significant. Moreover, once offshore, options for mitigation are limited or are met with high costs and/or significant delays. This paper elaborates on a model that was developed to predict the extent of pile run and velocity of pile and hammer during a drop fall. The model includes an approach that allows for assessment of effects from liquefaction during high velocity pile runs. A correction for strain rate dependent shear strength is also included. The model was validated by means of field observations of multiple controlled pile run events during construction of an offshore wind farm. More recently, pile run field observations from a monopile installation project became available to the authors, for which the model also proved satisfactory. The observed and analyzed pile run for this specific case is also reported. Use of the model provides insight into the extent of pile run and could provide a basis for assessment of mitigating measures. Sensitivity analyses were performed to determine which model parameters are key and what uncertainty bandwidth in pile run predictions should be taken into account.