Kriging meta-models for damage equivalent load assessment of idling offshore wind turbines

Research output: Contribution to journalArticleResearchpeer review

View graph of relations

Details

Original languageEnglish
Pages (from-to)3069-3089
Number of pages21
JournalWind Energy Science
Volume10
Issue number12
Publication statusPublished - 22 Dec 2025

Abstract

The use of meta-models (e.g. Kriging, artificial neural networks, and polynomial chaos expansion) as surrogate models of aeroelastic simulation models offers a good opportunity to perform lifetime calculations with a feasible computational effort. Meta-models for the approximation of fatigue loads, i.e. damage equivalent loads, of wind turbines in normal operation have been researched comprehensively in recent years. For offshore wind turbines in particular, however, downtimes, i.e. the times when wind turbines idle, also have a significant impact the lifetime. Currently, there are no meta-models of idling wind turbines available. However, it cannot simply be assumed that the findings from normal operation can be directly transferred to idling, as the structural behaviour differs significantly from normal operation due to the lack of aerodynamic damping and the resulting larger impact of the wave loads. For this reason, for the first time, the creation of meta-models, more precisely Kriging meta-models, for an idling offshore wind turbine is investigated comprehensively in this paper. The investigation of meta-modelling shows that for the approximation of the rotor blade root bending moments, two additional input parameters have to be considered in addition to the input parameters that are used for the creation of a meta-model for the same offshore wind turbine in normal operation. The comprehensive investigation of the Kriging meta-models shows that the meta-models trained with 2500 data points represent the simulation model with an acceptable approximation quality when choosing suitable Kriging settings.

Cite this

Kriging meta-models for damage equivalent load assessment of idling offshore wind turbines. / Schmidt, Franziska; Hübler, Clemens; Rolfes, Raimund.
In: Wind Energy Science, Vol. 10, No. 12, 22.12.2025, p. 3069-3089.

Research output: Contribution to journalArticleResearchpeer review

Schmidt F, Hübler C, Rolfes R. Kriging meta-models for damage equivalent load assessment of idling offshore wind turbines. Wind Energy Science. 2025 Dec 22;10(12):3069-3089. doi: 10.5194/wes-10-3069-2025
Schmidt, Franziska ; Hübler, Clemens ; Rolfes, Raimund. / Kriging meta-models for damage equivalent load assessment of idling offshore wind turbines. In: Wind Energy Science. 2025 ; Vol. 10, No. 12. pp. 3069-3089.
Download
@article{41dd343747534843afe9a98e9235e6bf,
title = "Kriging meta-models for damage equivalent load assessment of idling offshore wind turbines",
abstract = "The use of meta-models (e.g. Kriging, artificial neural networks, and polynomial chaos expansion) as surrogate models of aeroelastic simulation models offers a good opportunity to perform lifetime calculations with a feasible computational effort. Meta-models for the approximation of fatigue loads, i.e. damage equivalent loads, of wind turbines in normal operation have been researched comprehensively in recent years. For offshore wind turbines in particular, however, downtimes, i.e. the times when wind turbines idle, also have a significant impact the lifetime. Currently, there are no meta-models of idling wind turbines available. However, it cannot simply be assumed that the findings from normal operation can be directly transferred to idling, as the structural behaviour differs significantly from normal operation due to the lack of aerodynamic damping and the resulting larger impact of the wave loads. For this reason, for the first time, the creation of meta-models, more precisely Kriging meta-models, for an idling offshore wind turbine is investigated comprehensively in this paper. The investigation of meta-modelling shows that for the approximation of the rotor blade root bending moments, two additional input parameters have to be considered in addition to the input parameters that are used for the creation of a meta-model for the same offshore wind turbine in normal operation. The comprehensive investigation of the Kriging meta-models shows that the meta-models trained with 2500 data points represent the simulation model with an acceptable approximation quality when choosing suitable Kriging settings.",
author = "Franziska Schmidt and Clemens H{\"u}bler and Raimund Rolfes",
note = "Publisher Copyright: {\textcopyright} 2025 Franziska Schmidt et al.",
year = "2025",
month = dec,
day = "22",
doi = "10.5194/wes-10-3069-2025",
language = "English",
volume = "10",
pages = "3069--3089",
number = "12",

}

Download

TY - JOUR

T1 - Kriging meta-models for damage equivalent load assessment of idling offshore wind turbines

AU - Schmidt, Franziska

AU - Hübler, Clemens

AU - Rolfes, Raimund

N1 - Publisher Copyright: © 2025 Franziska Schmidt et al.

PY - 2025/12/22

Y1 - 2025/12/22

N2 - The use of meta-models (e.g. Kriging, artificial neural networks, and polynomial chaos expansion) as surrogate models of aeroelastic simulation models offers a good opportunity to perform lifetime calculations with a feasible computational effort. Meta-models for the approximation of fatigue loads, i.e. damage equivalent loads, of wind turbines in normal operation have been researched comprehensively in recent years. For offshore wind turbines in particular, however, downtimes, i.e. the times when wind turbines idle, also have a significant impact the lifetime. Currently, there are no meta-models of idling wind turbines available. However, it cannot simply be assumed that the findings from normal operation can be directly transferred to idling, as the structural behaviour differs significantly from normal operation due to the lack of aerodynamic damping and the resulting larger impact of the wave loads. For this reason, for the first time, the creation of meta-models, more precisely Kriging meta-models, for an idling offshore wind turbine is investigated comprehensively in this paper. The investigation of meta-modelling shows that for the approximation of the rotor blade root bending moments, two additional input parameters have to be considered in addition to the input parameters that are used for the creation of a meta-model for the same offshore wind turbine in normal operation. The comprehensive investigation of the Kriging meta-models shows that the meta-models trained with 2500 data points represent the simulation model with an acceptable approximation quality when choosing suitable Kriging settings.

AB - The use of meta-models (e.g. Kriging, artificial neural networks, and polynomial chaos expansion) as surrogate models of aeroelastic simulation models offers a good opportunity to perform lifetime calculations with a feasible computational effort. Meta-models for the approximation of fatigue loads, i.e. damage equivalent loads, of wind turbines in normal operation have been researched comprehensively in recent years. For offshore wind turbines in particular, however, downtimes, i.e. the times when wind turbines idle, also have a significant impact the lifetime. Currently, there are no meta-models of idling wind turbines available. However, it cannot simply be assumed that the findings from normal operation can be directly transferred to idling, as the structural behaviour differs significantly from normal operation due to the lack of aerodynamic damping and the resulting larger impact of the wave loads. For this reason, for the first time, the creation of meta-models, more precisely Kriging meta-models, for an idling offshore wind turbine is investigated comprehensively in this paper. The investigation of meta-modelling shows that for the approximation of the rotor blade root bending moments, two additional input parameters have to be considered in addition to the input parameters that are used for the creation of a meta-model for the same offshore wind turbine in normal operation. The comprehensive investigation of the Kriging meta-models shows that the meta-models trained with 2500 data points represent the simulation model with an acceptable approximation quality when choosing suitable Kriging settings.

UR - http://www.scopus.com/inward/record.url?scp=105026234640&partnerID=8YFLogxK

U2 - 10.5194/wes-10-3069-2025

DO - 10.5194/wes-10-3069-2025

M3 - Article

VL - 10

SP - 3069

EP - 3089

JO - Wind Energy Science

JF - Wind Energy Science

SN - 2366-7443

IS - 12

ER -

By the same author(s)