The majority of offshore wind turbines (OWTs) to date are installed on monopile (MP) foundations. Their load-deflection behavior is governed by the lateral soil-structure response. The commonly named API p-y method, often used for MP design, was originally developed to model the lateral response of long slender jacket piles based on test piles with penetration length to diameter ratios (L/D) of 34. These pile geometries are very dissimilar to the large-diameter MPs used in the offshore wind industry today, where L/D ratios of less than 3.5 are not unusual. The applicability of the API p-y method for the MP design under lateral loading is unclear, not only when extrapolating the method to much larger diameters and reduced L/D ratios, but also with regards to the soil-structure interaction mechanisms neglected by the model. Thus, a design methodology specifically targeting large-diameter MPs was developed under the PISA (Pile Soil Analysis) joint-industry project (e.g. Byrne et al. 2020 and Burd et al. 2020). This paper presents a comparison of the API p-y method for sand and the PISA general Dunkirk sand model (GDSM) (Burd et al. 2020) on the prediction of the response of laterally loaded MPs against field measurements from large-diameter onshore pile load tests (D = 4.3 m and L/D = 4.3). The pile testing was executed under the Carbon Trust and RWE Innogy led 'Vibration Pile Validation' (VIBRO) project, as part of the Offshore Wind Accelerator (OWA) program. The testing program took place at the Cuxhaven site in Germany and the site ground conditions consist of dense over-consolidated North-Sea sand. The comparison confirms that the PISA GDSM results in more accurate predictions of the lateral response of large-diameter piles and suggests that the API p-y method for sand may lead to non-conservative predictions.
Peter Duroska, Vasilis Avgerinos, Miguel Pacheco Andrade, Kris W. Andersen and Morten Albjerg Liingaard
Keywords: monopiles, API, PISA, offshore wind farms, geotechnics
Publisher: DFI