During the first phase of Supergen Wind project, the following achievements have been achieved:
| Identifying causes of turbine unreliability and low availability |
Reliability analysis of WTs has been established by the Consortium and through interaction with European partners, using public data, we have identified the key turbine subassemblies causing downtime. These data show that 75% of all failures are responsible for only 5% of the annual downtime and the remaining 25% of failures cause 95% of the downtime.
Figure 1. Using a presentation developed by ISET, showing the reliability and downtime from two surveys including more than 15000 turbine years. Extracted by the Consortium in collaboration with ISET & TU Delft.
|
| Identifying how and where condition monitoring can improve turbine availability |
The Consortium has reviewed commercially available WT condition monitoring packages and identified their strengths and weaknesses. This is being disseminated to the industry. In the light of the reliability and availability work above new condition monitoring techniques are being developed for WTs to improve effectiveness of early fault detection on the 25% of faults that cause the majority of downtime.
|
| Identifying control improvements to reduce WT loads improve reliability and reduce weight |
In large multi-MW WTs the production loads are a major contribution to lifetime equivalent fatigue loads on towers and blades. Loads on both can be reduced by improving WT controllers. Advanced solutions to the control design have reduced tower loads by 15-18% without compromising the WT performance. Similarly blade loads can be reduced by individual blade control, which has attracted interest from Industrial Partners.
Figure 2, Relative percentage lifetime load reduction collective pitch for various moments on different WT sizes using modified control
Figure 3, Time plot of the nodding motion moment, My, of a WT in stationary hub coordinates using modified control
|
| Producing high quality single and array models and measurements for turbine wakes |
Wind tunnel simulations have been set up with multiple, rotating, speed-controlled 5MW size model WTs in same-scale off- and on-shore neutral atmospheric boundary layers (ABL) and measurements made of all three velocity components, turbulence intensities, Reynolds stresses and other quantities. Measurements have also been made of wakes in a correctly scaled, stably stratified, off-shore ABL. Data gained is of direct interest to industry for improving currently inadequate methods and understanding. A wake/ABL rotor aerodynamic unsteady loading model has been developed.
Figure 4, Typical wake effects, which the Consortium are modelling, observed on a large offshore wind farm. (Source: Vestas)
Figure 5. Wind farm wake measurements being undertaken by the Consortium.
Figure 6, Consortium wind tunnel measurements of velocity 4D downwind of two scale-model WTs 2D apart.
Figure 7, Consortium CFD flow simulation around a forest clearing. High level wind vectors, above canopy, in blue and low level wind vectors in red. |
| Producing robust single and array Radar Cross Section (RCS) turbine models |
Computationally efficient models developed for studying interaction between improved lightning performance and reduced RCS, focused on the impact of WTs on marine navigational radar. Work has been well received by Industrial Partners. Derived and published fundamental new analysis on WT blade lightning strikes, which has fed into the latest revision of the standard IEC 61400-24.
Figure 8. Wind farm radar cross section (RCS) models developed by the Consortium.
|
| Investigating and testing new composite materials to improve blade strength and durability |
New composite interlayer reinforcements (hybrid veils) developed that increase blade structure fatigue life by half an order of magnitude, while simultaneously increasing toughness threefold in bonded connections. This has improved damage tolerance without accompanying disruption in blade manufacturing processes. New structural fibres also evaluated that confirm order of magnitude improvements in blade spar cap fatigue life.
|
| Developing materials to reconcile the simultaneous reduction of RCS and improvement of lightning performance |
Novel layers have been introduced into the composite that incorporate metal coated fibres in a non-woven layer. This has reduced RCS and increased electrical conductivity in the laminate without causing any associated reduction in blade strength.
Figure 9, New blade interlayer materials, selected by Consortium, increase toughness leading to improved fatigue performance.
|
| Developing a parametric blade model for the assessment of innovative materials, condition monitoring effectiveness, and “smart” blade performance |
A parametric blade model, comprising a script-driven pre-processor to the Abaqus finite element program, has been developed. The script enables parametric studies of the positioning of key structures, the application of true aerodynamic loading, as blade surface pressure, and the assessment of the importance of improvements in material properties. An extension to the model has been made to aid interpretation of results from the NAREC biaxial resonant load fixture. Figure 8.
Figure 10. WT parametric blade model developed by the Consortium.
|
| Modelling and testing the prediction of scour around offshore turbine foundations |
A detailed understanding of the relationships between the hydrodynamics, bed profile, and pile geometry is required for scour mitigation. Detailed flow measurements were taken from a series of tank experiments and the data is being analysed. A new, efficient numerical model of current induce scour was developed. Special attention was given to choice of turbulence model which was validated against published data.
Figure 11, Photo of Consortium test tank showing a circular monopile and pressure sensors.
Figure 12, Mean centreline component of stream velocity around turbine foundation in Fig 9 from Consortium hydrodynamic model showing comparison with experimental data. |
| Analysing blade stress distribution |
Stress distribution in Finite Element Analysis (Figure 13) shows the stresses are mainly concentrated around the intersection. When the interlaminar fracture toughness of T-joint reaches a specific level, the T-joint samples would prefer to fail with laminate fracture at intersection area rather than propagate the delamination cracks ( Figure 14 ).
Figure 13, FE analysis of stress distribution on T-joint sample under tensile loading
Figure 14, Laminate fractures instead of propagating delamination cracks under tensile loading |
