A comparison study on jacket substructures for offshore wind turbines based on optimization

Abstract

Abstract. The structural optimization problem of jacket substructures for offshore wind turbines is commonly considered as a pure tube dimensioning problem with given topology, minimizing the entire mass of the structure. However, this approach goes along with the assumption that the given topology is fixed in any case. The present work contributes to the improvement of the state of the art by utilizing more detailed models for geometry, costs, and structural design code checks. They are assembled in an optimization scheme, in order to consider the jacket optimization problem from a different point of view 5 that is closer to practical applications. The objective function is replaced by a sum term comprising several cost terms. To address the issue of high demand of numerical capacity, a machine learning approach based on Gaussian process regression is applied to reduce numerical cost and enhance the number of considered design load cases. The proposed approach is meant to provide decision guidance in the first phase of wind farm planning. A numerical example for a NREL 5 MW turbine under FINO3 environmental conditions is computed by two effective optimization methods (sequential quadratic programming and 10 an interior-point method), allowing for the estimation of characteristic design variables of a jacket substructure. In order to resolve the mixed-integer problem formulation, multiple subproblems with fixed integer design variables are solved. The approach shows reasonable and promising results, useful both for further research and technical applications.

Gaussian process regression for fatigue reliability analysis of offshore wind turbines

“…The lifetime damage values show very high accuracy compared to the scatter in Figure6; this is partly due to an averaging effect when considering all the environmental conditions (in fact, as 47.6% of the samples are above the 1:1 trendline and 52.4% are below it). These findings are in very good agreement with those of Slot et al[51], who used a Matern 3/2 kernel to represent fatigue loads on an onshore wind turbine, and Häfele et al[25] who found a Matern 5/2 kernel to be best for representing joint fatigue loads for a OWT on jacket substructure.…”

“…A set of 2,000 load cases were reduced by a factor of 10, with a reported error of less than 1% in terms of DEL [24]. Häfele et al [25] found that similar reductions were possible for an OWT located at a different North Sea site, suggesting that the number of load cases can be reliably reduced. However, the computational burden for FLS assessment remains large as the full set of load cases need evaluated to calibrate the reduced set.…”

Dynamic analysis of 10 MW offshore wind turbines with different support structures subjected to earthquake loadings

“…While the monopile and tripod adopts a transition structure that connects them to the tower instead of a direct connection to the platform, a jacket is often a combined connection as a monolithic unit with the tower. An innovative jacket transition structure for 10 MW wind turbines, named the optimized transition piece (OTP) -Figure 4, investigated in this paper was proposed in the INNWIND project [49]. The transition piece is designed as X-braces, X-joints and K-joints structure.…”

Reliability-based design optimization of offshore wind turbine support structures using analytical sensitivities and factorized uncertainty modeling

“…Overall, this was found to be the most robust for the regression problem in this study, especially when considering repeated regression for additional iterations of the outer loop in Algorithm 1. The Matérn class of kernels was also used for OWT support structures in Häfele et al (2019). We also note here that in order to simplify the simultaneous regression with respect to all of the three parameters in θ q , these parameters were 10 input to the fitting problem in such a way that the surrogate model became co-monotonic in every variable (an increase in one or more variables giving always an increase in the output).…”

“…While not strictly in the same category, studies where the fatigue calculation is based on crack propagation models and an assumption that the stress cycles follow a Weibull distribution, allowing exact limit state though because of the added conceptual complications these methods have yet to be applied in studies considering a more realistic and comprehensive set of loading conditions. A study founded on gradient-based optimization that does consider a more comprehensive set of loading conditions, but does not utilize analytical sensitivities, was performed by Häfele et al (2019). They used a Gaussian process surrogate model to simplify the response, thus making the analysis computationally feasible.…”