Accurate prediction of long-term 'characteristic' loads associated with an ultimate limit state for design of a 5-MW bottomsupported offshore wind turbine is the focus of this study. Specifically, we focus on predicting the long-term fore-aft tower bending moment at the mudline and the out-of-plane bending moment at the blade root of a monopile-supported shallow-water offshore wind turbine. We employ alternative probabilistic predictions of long-term loads using inverse reliability procedures in establishing the characteristic loads for design. Because load variability depends on the environmental conditions (defining the wind speed and wave height), we show that long-term predictions that explicitly account for such load variability are more accurate, especially for environmental states associated with above-rated wind speeds and associated wave heights.

The Geotechnics Sub-Committee of the American Society of Civil Engineers (ASCE) Coasts, Oceans, Ports, and Rivers Institute (COPRI) Marine Renewable Energy (MRE) Committee is preparing a guide document for marine renewable energy foundations. That guide would use standard design codes for fixed foundations and mooring anchors in API RP 2GEO and DNV.The static method of computing axial pile capacity described in API RP 2GEO (2011) is generally used to compute ultimate compressive and tensile capacities of pipe piles driven to a given penetration. Lateral soil resistance -pile deflection (p-y) data for clays and sands are usually developed using procedures proposed by Matlock (1970) andMurchison (1983), respectively, and outlined in API RP 2GEO (2011). Marine energy foundations are unique in several ways. Axial pile capacity computations are usually based on a reasonable lower bound, in contrast to the soil resistance to driving, which is based on a reasonable upper bound. For structures supporting wind turbines, however, underestimating (or overestimating) the soil stiffness could require a change in turbine operation and a loss of power production. Although the classical API method is recognized as an appropriately conservative design method for offshore pile foundations, a prediction method is more well suited for structures supporting wind turbines, such as the CPT-based methods for predicting pile capacity in granular soils presented in API RP 2GEO (2011). If a prediction method is used to compute the soil resistance to driving, the evaluation of pile drivability may be overly conservative. Ageing in both clay and sand should also be taken into account. Wind turbines are often supported on large diameter monopiles. The applicability of the p-y data for such large diameter piles needs to be verified. Finally, marine renewable energy generated by in-stream hydrokinetics, ocean thermal energy conversion, and wave energy converters may be floating devices usually anchored to the seafloor. There are uncertainties in the design and installation of these anchors, which become critical for large sustained tensile loads that may degrade due to creep and cyclic loading.

The objective of this paper is to address the applicability of using API RP 2GEO (2011) for the design of wind turbine monopile foundations in normally to moderately overconsolidated clays. The study involved three-dimensional numerical modeling using the finite-element method, one-g laboratory model testing, and analysis of field test results. The following conclusions concerning the use of Matlock (1970) soft clay p-y curves for the design of large-diameter monopile foundations are drawn: Numerical modeling and model-scale testing with rigid piles of different diameters indicate that the form of the Matlock (1970) p-y curves, in which the lateral displacement is normalized by pile diameter and lateral soil resistance is normalized by the ultimate resistance, appropriately captures the effect of pile diameter.Field and model testing indicate that the Matlock (1970) p-y models consistently overestimate the lateral displacements at the pile head when used to analyze laterally loaded piles in normally to moderately overconsolidated clays.An approximate version of the Jeanjean (2009) p-y model, in which the Matlock (1970) p-y curves are scaled by p-multipliers calculated at various depths, generally provides a reasonable match to measured lateral displacements at the pile head when a relatively large strain at one-half the undrained shear strength is assumed, i.e., ?50 = 0.02. This result applies both to small scale model tests in kaolinite and large-scale field tests in high-plasticity clay.Model tests show that cyclic loading causes the stiffness of the lateral pile-soil response to degrade by 20 to 30 percent. The amount of degradation is dependent on the displacement amplitude and the number of cycles. All of the degradation happens within 100 cycles, after which the stiffness is reasonably constant.Model tests show that the ultimate lateral capacity of the pile is not significantly affected by the previous cyclic loading.