◀ Back to projects overview P201101-008-TUD

 

Tools: Integrated design tools development for large number of varied structural configurations

Public summary

Offshore wind turbines are exposed to continuous excitation from wind and waves. These excitations are time-varying, both on the short scale – waves and turbulence – as on the longer scale, as a result of the changing weather conditions. The loads on the wind turbine, as a result of this wind and wave excitation, are not easily determined, because they depend on the motion of the turbine. This implies that the estimation of the wind loads should account for the structural motion resulting from  the wave loads, and vice versa. An integrated model – a model which combines the aerodynamic and hydrodynamic interactions – is required to account for the full dynamics of the system, in order to develop cost effective offshore wind turbine designs. The current project aimed at the development of such an integrated model, particularly for the design of the support structure.

The importance of a correct representation of the dynamics of offshore wind turbines is related to the prediction of the fatigue damage throughout the structural life-time. The stress variations, resulting from the time-dependent forcing, induce micro-cracks in the structural material, which on the long term may lead to structural failure. To prevent such a failure within the structural life-time, design calculations are required to prove a sufficient structural resistance. The difficulty here is that the precise loads during the design life-time of 20 years are not known beforehand. Therefore, many time simulations for prescribed load scenarios need to be performed to make sure that the expected fatigue damage does not exceed the maximum damage allowed. The developed integrated model aims in particular on the estimation of the design life-time of the support structure with respect to fatigue damage, allowing for a quicker assessment of the multitude of design cases that needs to be
considered for this evaluation.

A number of software tools exists, with which the aerodynamic interaction of a wind turbine can be calculated. To mention a few: BLADED, BHawC, FAST, FLEX5. These tools allow for a detailed analysis of the wind loads on the structure. In addition, some of these packages allow for hydrodynamic analyses as well, or alternatively, can be coupled with a hydrodynamic package. As a consequence of the high level of detail of these packages, the calculation time required for the fatigue damage verification is very high. With the application of these software tools, very limited space for design optimization is available throughout the design stage of an offshore wind turbine. Generally, only one design can be elaborated on for a complete wind farm, which covers many square kilometers. The chance to develop a couple of different designs, in order to optimize within a wind farm, is often missed. On top of this, the design of offshore wind turbines is traditionally divided over two parties: the turbine manufacturer and the offshore contractor. These parties design each a part of the offshore wind turbine – either the turbine or the foundation, applying their own design model each. In practise, this approach requires the sharing of the predicted forces at the connection between the turbine and the foundation – the so-called interface forces. It is very likely that this design procedure results in suboptimal offshore wind turbine designs, since none of the parties can develop a design that is optimal for both the turbine and the support structure.

Considering specifically the support structure of offshore wind turbines – consisting of the foundation and the tower – simplified models can potentially speed-up the preliminary design stage, allowing for the elaboration of multiple designs and iterations of each design in itself. The benefits of such a design approach can be expressed in terms of a cost reduction resulting from the opportunity to use the saved design time for the optimization of multiple designs within a wind farm, leading to a reduction in the required total amount of primary steel for all the support structure within the park. Such a simplified model should be integrated, i.e., combine the turbine and the foundation in one model, without the need to share detailed design information between different parties. Recent research in the optimization of the design process of offshore wind support structures focused
on the integrated modelling of the turbines, accounting for a detailed rotor and a simplified support structure representation. In addition to this, a separate detailed support structure model is still required for the support structure design. With the application of the integrated model the interface forces at the connection between the turbine and the support structure are computed, with which subsequently the support structure can be designed on the basis of the second model. Even though it is recognized that this approach allows for the isolated design of the support structure, an approach which potentially speeds-up the design process, the large number of time simulations still need to be performed with the detailed aerodynamic tool. Because of the high computational costs, design iterations are preferably avoided.

It can be concluded that the need exists for a simple integrated model, which allows for the relatively quick assessment of a large number of load cases, in order to derive preliminary support structure designs. Since the design focusses on the support structure, it seems obvious that the rotor-nacelle assembly representation should be simplified. Even more, because the detailed modelling of the aerodynamic interaction is mostly responsible for the high computational costs. The rotor-nacelle assembly consists of the nacelle, within which the mechanical energy is transformed into electrical energy and a generally three-bladed rotor. The blades of the rotor are long and slender flexible elements, which are shaped such to optimize the aero-elastic performance. Therefore, their crosssectional shape varies along the length, as well as the twist angle. Modern machines are pitchregulated, implying that both the pitch angle of the blades and the rotational speed can be adjusted for optimal energy production, or load reduction. A simplified rotor model should account for the mass of the nacelle and the blades. Moreover, the control parameters – pitch angle and rotational speed – should be accounted for explicitly. It should be noted, however, that for a reduced rotor representation the validity under extreme weather conditions decreases and requires validation with a detailed integrated model.

A third important aspect of the turbine rotor is the aerodynamic damping. In fore-aft motion, the rotor experiences a resistance from the surrounding air. This resistance is such, that the motion resulting from wave actions is significantly reduced. In fore-aft direction, this damping is widely recognized, while in the side-to-side direction its effect is often neglected. Despite the low damping value in side-to-side direction, its importance in case of wind-wave misalignments should not be underestimated. The reduced rotor representation for the integrated preliminary design of offshore wind support structures can only be successful if the aerodynamic damping is accounted for correctly. 

The development of the integrated preliminary design model for offshore wind support structures, including a reduced rotor representation, follows four steps:

1. The dynamic assessment of the aerodynamic interaction of a single rotating blade.
2. The analysis of the aerodynamic damping of a full rotor in both fore-aft and side-to-side direction. 
3. The definition of the fully integrated model, including the reduced rotor representation, a hydrodynamic interaction model and a soil-structure interaction model.
4. The validation of the integrated preliminary design model with an existing detailed software tool

The dynamic assessment of the aerodynamic interaction of a single rotating blade (step 1) is required to determine the important aspects with respect to the reduced rotor representation. To this end, a model is developed of a flexible rotating blade, accounting for the twist and the pitch, and the rotational velocity. Multiple aerodynamic interaction models exist, some of which describing the separation of the air flow round the blade and the subsequent load effects of the resulting vortices, while others assume a permanently attached flow. In terms of computational time, the latter models are preferred, and therefore the limits of application of such models for the single rotating blade are assessed.

Having defined the applicable aerodynamic interaction model and its limitations, the dynamic characteristics of a full rotor are assessed. Explicit expressions for the fore-aft and side-to-side aerodynamic damping are derived, accounting for the rotational velocity and the pitch of the blades (step 2). The analysis includes possible deviations of separate blades, with respect to mass and geometry. The relevance of the side-to-side damping is demonstrated, particularly for higher wind velocities. Moreover, the analysis addresses the eccentricity of the rotor with respect to the tower axis, which introduces a coupling between the side-to-side motion and the yaw motion of the turbine, resulting in an increase of the effective side-to-side damping.

On the basis of step 2, the reduced rotor representation can be developed, which allows for the definition of the fully integrated model (step 3). This model includes the hydrodynamic and soilstructure interaction. Special attention is paid to the soil model, for which use is made of a novel approach, developed at Delft University of Technology, which enables a simple and accurate description of the complicated 3D soil behavior. The final part of the project concerns the validation of the model. To this end, simulations are performed with both the simple integrated model and BLADED, a detailed commercial aero-elastic tool. The comparison is done with respect to both computational time and accuracy of the structural response predictions. With respect to the latter, the results obtained with BLADED are taken as a reference. Further validation should be done on the basis of measurement data from an actual wind farm. This data will be obtained from currently planned measurement campaigns from newly constructed offshore wind farms. Assessment of the impact of the tool on the design process requires the application of the tool during the design loop of an offshore wind farm. 

The current project is part of FLOW R&D theme 2: Support Structures and contributes to the R&D line 2.2 Design and Certification. The proposed simplified integrated design model potentially allows for quicker design iterations and corresponding design optimizations. In addition, the model allows for an integrated design approach, focussing particularly on the support structure. In this respect, the project can be seen to contribute to R&D line 2.4 Integration of support structure within the faroffshore wind farm too.

With the application of the integrated preliminary design tool with the reduced rotor representation, the added value of the preliminary phase can be increased. This potentially reduces risk of suboptimal designs resulting from the separate design approached by the turbine manufacturer and the offshore contractor. A more elaborate preliminary design phase allows for a quicker assessment of the feasibility of an offshore wind project. Moreover, multiple design variations can be considered in a shorter period of time. Eventually, it is expected that a more elaborate preliminary design phase allows for quicker designs and more time for design optimizations and corresponding cost reductions. In addition, a more accurate implementation of the aerodynamic damping allows for more costeffective designs too.

The resulting cost reduction can best be expressed in terms of a reduced amount of primary steel required for the turbine tower and the support structure. The overall mass reduction of both tower and support structure is currently estimated as 10%. This reduction is assumed for both the monopile and the lattice structure foundation. It is expected that the cost reduction for a lattice foundation will differ from the reduction for a monopile foundation, because the cost-breakdown differs. The main cost reduction, however, results from the opportunity to optimize the support structures better, as a result of the speed-up of the preliminary design phase. On this basis, there is no reason to assume a different cost reduction for the lattice structure foundation than for the monopile foundation.

With the help of the FLOW cost model, this weight reduction is translated in an overall reduction of the levelized cost of energy for offshore wind. It is shown that the expected reduction of the levelized cost of energy for offshore wind, resulting from the application of the preliminary integrated design model, is between 1.0 and 1.5%.