The potential lifetime of wind turbine components is usually not fully utilized as the site conditions are less severe than assumed in the turbine design. Operators of wind farms can make use of the excess fatigue budget to increase the energy yield and thus decrease the levelized cost of energy (LCoE). To achieve this, the lifetime of the turbine can be extended until the fatigue budget is exhausted. Alternatively, a rotor blade extension (RBE) is an option to increase the energy yield of a wind turbine. An RBE increases the blade length and thus the swept area and the energy yield. An RBE also increases the loads on the turbine, however. Higher fatigue loads in turn reduce the fatigue budget of a turbine. This study investigates whether the use of an RBE is advantageous compared with a sole lifetime extension (LTE). As the use case, a commercial 1.5 MW turbine located in northern Germany was investigated. Aeroservoelastic multibody load simulations and simplified static load simulations were verified with each other. These simulations revealed the loads to determine the fatigue budget of the turbine components. Since the blade became the critical component when a certain RBE length was exceeded, the blade was subjected to a structural fatigue analysis. The fatigue analysis focused on the trailing-edge bond line which became critical when lead–lag loads increased with blade length. Finally, the energy production gains due to LTE and RBE were assessed. For the use case turbine, this study revealed an LTE of 8.7 years after a design life of 20 years with an additional energy yield of 43.5 %. Moreover, the extension of the 34 m blade with an RBE length of 0.8 m further increased the yield by 2.3 %.
Wind turbine operators are faced with a decision of how to handle their wind turbines once their certified lifetime has been reached.
Common options are dismantling, repowering, or a lifetime extension (LTE) of the turbine
Older turbines were designed according to general wind conditions that were reflected by wind turbine classes according to
This study investigates the feasibility of a rotor blade extension (RBE) as a retrofit solution installed early in the turbine lifetime to supplement an LTE. Both options are aimed at increasing the total energy production of the turbine during its lifetime and hence at decreasing the levelized cost of energy (LCoE).
An RBE increases the energy production by increasing the swept area of the turbine.
The loads on the rotor blade and the turbine are increased as well, however.
Load mitigation strategies need to be employed to limit the ultimate loads on the turbine
The driving loads for the assessment of an LTE and an RBE are the fatigue loads.
The lead–lag fatigue loads in particular are mostly generated by gravity loads on the blade during each revolution.
Hence, increased lead–lag loads are inevitable as a consequence of the blade tip mass added by an RBE.
The increase in lead–lag fatigue loads can be critical for the initiation of tunneling cracks in the adhesive bond line at the trailing edge, which can be initiated early in the turbine lifetime and may propagate into the blade structure
The design loads were obtained from a simplified simulation that takes into account static aerodynamic flapwise mean loads superimposed with alternating lead–lag gravity loads.
The results of the static simulation were verified with the results of an aeroelastic multibody simulation.
The use case investigated for this study was a 1.5 MW Südwind S70 wind turbine.
The turbine was installed in 2003 in the Bremervörde-Iselersheim wind farm located in northern Germany
In analogy with the concept of extending already-operating blades with an RBE,
To the authors' knowledge, this work is the first time the impact of an RBE on the total lifetime including LTE with respect to the maximization in energy yield has been presented in the internationally available literature published in English.
This paper is structured as follows:
Sect.
A 1.5 MW turbine was selected for the use case scenario in this work.
An RBE was designed for this use case turbine.
As per the design requirement the blade structure was not allowed to be altered; i.e., the blade tip was not allowed to be cut up.
Therefore, the concept of extending already-operating blades at the tip with an overlapping RBE was investigated.
The concept (Fig.
Design concept of a rotor blade extension.
The RBE consists of two half shells that are pulled over an existing blade and fixed to it with adhesive. A rib located at the original blade tip is used as a second support. A web is added to the RBE to transfer the shear loading. The lightning protection system of the original blade passes through the blade tip of the root segment into the RBE. A lightning receptor is required at the tip of the RBE. Moreover, an outlet hole for water was considered in the RBE. This design is beneficial compared with designs containing support rods because no holes need to be drilled through the skin of the blade structure. Moreover, the RBE could potentially be dismantled with relatively little effort. Furthermore, the root blade segment remains unmodified.
The RBE is mounted on-site on the turbine without dismantling the entire rotor (Fig.
Siemens Bonus 1.5 MW turbine
In this work, the length of the RBE was parametrically increased to investigate its impact on AEP, fatigue budget, turbine dynamics, and tower clearance.
The LTE assessment of all wind turbine components was performed by comparing design loads to site-specific loads. This assessment used the loads from an aeroservoelastic load simulation. In addition, a simplified static load simulation was performed to evaluate the loads on the trailing edge of the blade. The two types of load simulations were conducted for different lengths of the RBE. A structural model of the rotor blade was used to evaluate the fatigue damage in the blade.
The procedure of the static load simulation conducted in this work has been described by
The lead–lag fatigue load cycles are calculated on the basis of the rotor revolutions in the design lifetime of the blade.
The number of revolutions
The rotation frequency is defined as a function of the wind speed:
The lead–lag revolutions of the SSP34 during its lifetime were estimated for the site conditions at Bremervörde-Iselersheim (Fig.
Probability density of wind speed of design condition and at actual site.
Design parameters of 1.5 MW turbine.
The simplified load calculation was performed to calculate the loads on the trailing-edge bond line. In the simplified load calculation, static mean flapwise aerodynamic loads for 12 wind speed bins were superposed with the alternating lead–lag gravity loads due to the rotor revolution. The blade pitch angle was taken into account in the calculations.
A beam model implemented in APDL (Ansys Parametric Design Language;
At operating wind speeds in the partial load regime of the turbine, a pitch angle working point of
The use case turbine was modeled in the multibody dynamics simulation software MSC ADAMS (Automated Dynamic Analysis of Mechanical Systems;
For the rotor blades, the distribution of mass and bending and torsional stiffness along the blade span was considered using the shear center, center of gravity, and elastic center for each cross section.
From the detailed FE model, modal bodies for use in ADAMS were derived using constraint and fixed boundary modes in conjunction with the Craig–Bampton method
For the tower structure, a stiff-stiff tower design was modeled.
The first eigenfrequency of the tower was tuned to 0.38 Hz, corresponding to 1.2 times the rated rotor speed
As the load case for the fatigue limit state, DLC 1.2 according to IEC 61400-1
The fatigue budget of relevant turbine components was evaluated at the different wind sites on the one hand and for different RBE lengths on the other. Hence, a stress-based damage analysis was conducted for the rotor blade structure and a damage-equivalent-load-based analysis for the other turbine components.
An Euler–Bernoulli
On the basis of the formulation by
The individual materials of a rotor blade, i.e., adhesive, resin, and glass fiber, can be considered to be isotropic.
Assuming that a symmetric constant life diagram
Assuming that the damage accumulation according to
The method of load and resistance factor design
The blade material properties for this study were obtained experimentally and have been presented by
Furthermore, for the static load simulation case, a load factor of
For the integrity analysis according to
From the static load simulation, mean and amplitude loads were extracted at each wind speed bin.
These load spectra were used to calculate stress spectra, the damage Eq. (
From the aeroelastic load simulation, time series of each blade cross section were extracted at each wind speed bin.
Initially, the beam model was solved for unit load cases.
Thereafter, the load time series were used to determine strain time series along the trailing edge.
A rainflow-counting algorithm
The fatigue budget analysis of the relevant components, i.e., blade root, blade bolts, hub, shaft, main frame, tower top and bottom, and foundation, was conducted on the basis of a damage equivalent load amplitude (DEL) as proposed by
The loads for the design and site situation were obtained with an aeroelastic simulation, as described above.
After rainflow counting the moment histories, the DEL for each turbine component was obtained according to
Moment history at
Strain history at
The results of the two simulation approaches, i.e., the static and the aeroelastic load simulation, were benchmarked with each other.
First, the bending moment history of the aeroelastic simulation was compared with the minimum and maximum moments of the static simulation. Figure
Markov matrix of a 600 s aeroelastic simulation at
Markov matrix of a 600 s static simulation at
Relative damage from aeroelastic simulation.
Second, the strain history in the trailing edge is illustrated in Fig.
Relative damage from static simulation.
Moreover, the rainflow-counted strain history of the aeroelastic and static simulation is presented as a Markov matrix in Figs.
Third, the relative damage impact of each wind speed
Fourth, the stress exposure along the blade span is illustrated in Fig.
Fatigue stress exposure along the trailing-edge bond line obtained with the two simulations.
Campbell diagrams for the original turbine and the turbine with an RBE.
When the RBE length is increased, the serviceability limit state of the turbine needs to be considered. The blade tip-to-tower clearance must remain large enough, and resonance as a result of changing eigenfrequencies must be avoided.
The Campbell diagram (Fig.
The tip-to-tower clearance
Blade deflection and tip-to-tower clearance in meters.
Neither an extrapolation to the 50-year value nor the extreme turbulence model is applied as would be required by
The static simulation revealed the loads for the fatigue analysis of the blade.
The fatigue stress exposure is plotted along the blade span for different RBE lengths (Fig.
Stress exposure along the trailing-edge bond line for different blade extension lengths obtained with the static simulation.
Considering the blade without RBE, Eq. (
Fatigue budget for the critical turbine component
The fatigue budget or remaining lifetime of all other turbine components (Table
Lifetime extension of components in years.
The red, orange, and green bars in Fig.
The red bar represents a scenario in which no RBE is considered and the fatigue budget of the turbine, e.g., the shaft
The green bar represents a scenario in which the fatigue budget of any component other than the blade limits the lifetime of the turbine:
The orange bar represents the actual scenario of our use case turbine.
Herein, the fatigue budget of the shaft limits the turbine life to
The latter scenario assumes that the blade with RBE does not reduce the fatigue budget of the shaft due to increased loads.
As discussed before, the difference in fatigue damage in the shaft as a result of increased RBE length is negligible, so
The energy production as a function of the RBE length is expressed as
The change in energy production as a function of the RBE length EP
Relative energy production increase as function of blade length. Three scenarios shown.
Again, red (here a red dot) indicates a scenario in which the blade limits the fatigue budget of the turbine.
For any RBE length, the energy production loss caused by the smaller turbine fatigue budget outweighs the gains caused by an increase in the swept rotor area (blue curve).
When only an LTE of
The orange line represents our use case.
In this scenario, the RBE length can be increased until
The green curve represents the scenario in which the fatigue budget of any component other than the blade limits the lifetime of the turbine:
The static simulation was verified for analyses of the trailing edge in the inboard blade region at up to 40 % of the blade length when an additional load factor of
A relatively detailed damage analysis of turbine components, as conducted for the blade structure in this study, is not possible in most cases since design information is not available because it is confidential or it simply does not exist anymore. Therefore, the reverse engineering of the turbine and the method of relative damage equivalent load comparison is the obvious approach for the LTE calculation.
The relative DEL comparison is expected to yield conservative results. However, if more design information about the geometry and materials, especially for critical turbine components such as the blade bolts, blade root, and shaft, were available, a more reliable assessment of the lifetime extension would be possible.
A lifetime extension is feasible every time the site conditions are subjected to a lower average wind speed than the design conditions. The extended lifetime depends on the fatigue budget of the critical turbine component. When economical, it is feasible to replace critical components, e.g., blade bolts, in order to extend the lifetime.
Considering our use case, the energy production gain solely with an LTE is 43.5 % (Fig.
Considering our use case where any component other than the blade is driving the LTE, the RBE increases the energy production by 2.3 % on top of the energy gain resulting from an LTE (43.5 %), as explained above.
In this case, the RBE can be increased up to a length at which the blade itself becomes the critical turbine component for the turbine lifetime.
This length is interpreted as the optimum RBE length.
Above the optimum length, fatigue budget losses outweigh any energy production gains.
Our use case scenario shows that, for a 34 m blade, an RBE length of 0.8 m is most advantageous (Fig.
The optimum RBE length obtained is within the boundaries defined by a serviceability limit state analysis which yielded a maximum RBE length of
We point out that the additional yield due to an RBE needs to cover the costs for the design, manufacturing, and installation of the RBE.
The static simulation was verified with an aeroelastic simulation which is applicable to structural analyses of the trailing edge in the inboard blade region at up to 40 % of the blade length.
A lifetime extension is feasible every time the site conditions are subjected to a lower average wind speed than the design conditions. The extended lifetime depends on the fatigue budget of the critical turbine component.
Considering our use case, the energy production gain solely with an LTE is 43.5 %. On top of the energy gain produced by an LTE, the RBE further increases the energy production by 2.3 %. For the 1.5 MW turbine in our use case, an RBE length of 0.8 m was most advantageous for the 34 m blade. The length obtained was within the limits for turbine resonance and tower clearance.
When the mean stress
The data presented in the figures are available at
MR conducted the static load simulations and the fatigue analysis of the rotor blade. MS modeled the use case turbine and its controller, conducted the aeroelastic load simulations, and carried out the fatigue analysis of the turbine components. The two authors assessed the rotor blade extension retrofit and the lifetime extension of the turbine together.
The authors declare that they have no conflict of interest.
This article is part of the special issue “Wind Energy Science Conference 2019”. It is a result of the Wind Energy Science Conference 2019, Cork, Ireland, 17–20 June 2019.
We acknowledge the support of P. E. Concepts GmbH. Moreover, we would like to thank SSP Technology A/S for providing the geometry, the laminate plan, and the material properties for the wind turbine blade model for this research.
This research has been supported by the Bundesministerium für Wirtschaft und Energie within the SmartBlades2 project (grant no. 0324032B).
This paper was edited by Raimund Rolfes and reviewed by two anonymous referees.