Achieving the 20 MW Wind Turbine

The industry is taking two parallel pathways towards cost reduction: first, incremental innovation, looking at cost reductions through economies of scale resulting from increased market volumes of mainstream products, with a continuous improvement of manufacturing and installation methods and products; secondly, breakthrough innovation, focused on creating innovative products, including significantly upscaled dedicated offshore turbines, to be considered as new products.

Funded under the European Union's Sixth Framework Programme (FP6), the UpWind project has been exploring both pathways. The UpWind project initiators realised that wind technology disciplines were rather fragmented, in that no integrated verified design methods were available; that essential knowledge was still missing in high priority areas, for example in external loads; that measuring equipment was still not accurate or fast enough; and that external factors were not taken into consideration in minimising cost of energy (grid connection, foundations, wind farm interaction and so on).

Optimum Technology?

Rather than pursue one single 'optimum' technology, UpWind explored various high-potential solutions and integrated them with a view to the potential reduction of cost of energy. An optimised wind turbine is the outcome of a complex function combining requirements in terms of efficiency of electricity production, reliability, access, transport and storage, installation, visibility, support to the electricity network, noise emission, cost, and so forth. UpWind's focus was the wind turbine as the essential component of a wind power plant. Thus external conditions were only investigated if the results were needed to optimise the turbine configuration.

In Search of the 20 MW Wind Turbine

UpWind established that 'the same, but bigger' would simply not work. Starting with a reference 5 MW wind turbine it initially extrapolated this reference design to upscale it to 10 MW, and then on to 20 MW.

This virtual 20 MW design was unanimously assessed as almost impossible to manufacture, and uneconomic. It would weigh 880 tonnes on top of a tower making it impossible to store today at a standard dockside, or to install offshore with the current installation vessels and cranes.

The support structures able to carry such mass, placed at 153 metres height, are not possible to mass manufacture today. The blade length would exceed 120 metres, making it the world's largest ever manufactured composite element, which cannot be produced as a single piece with today's technologies. The blade wall thickness would also exceed 30 cm, which puts constraints on the heating of inner material core during the manufacturing process, and the blade length would also require new types of fibres in order to resist the loads.

UpWind's 'Lighthouse' Approach to Methodology

This 20 MW concept was drawn up to provide values and behaviour that could be a starting point for optimisation. UpWind then developed innovations to enable this basic design to be significantly improved, and therefore enable development of a potentially economically sound design.

A means of doing this was adopting the lighthouse - a virtual concept design of a wind turbine in which promising innovations, either mature or embryonic, are incorporated. The lighthouse is not a pre-design of a wind turbine that will actually be realised, but a concept from which ideas can be drawn by industry for product development. One example is a blade made from thermoplastic materials, incorporating distributed blade control, including a control system, the input of which is partly fed by LIDARs.

The 20 MW Innovative Turbine

While the 20 MW design was unfeasible due to issues of weight and corresponding loads, the future large-scale wind turbine system drawn up by the UpWind project is smart, reliable, accessible, efficient and lightweight.

It analysed wind turbine materials, optimising the micro-structure of blade materials to develop stronger and lighter blades. Fatigue loading also needed to be reduced so that longer and lighter blades may be built. And the aerodynamic and aeroelastic qualities of the models were significantly improved. Significant knowledge was gained on load mitigation and noise modelling.

UpWind demonstrated that advanced blade designs could alleviate loads by 10%, by using more flexible materials and fore-bending the blades in the second work programme (WP). After reducing fatigue loads and applying materials with a lower mass to strength ratio, a third essential step is needed: the application of distributed aerodynamic blade control, requiring advanced blade concepts with integrated control features and aerodynamic devices. Fatigue loads could be reduced by 20%-40% (WP2).

Various devices can be utilised to achieve this, such as trailing edge flaps, (continuous) camber control, synthetic jets, micro tabs, or flexible, controllable blade root coupling. Within UpWind, prototypes of adapting trailing edges, based on piezo-electrically deformable materials and SMA (shape memory alloys) were demonstrated (WP1B.3).

However, the control system only works if both hardware and software are incorporated in the blade design. Thus advanced modelling and control algorithms need to be developed and applied. This was investigated in WP1B3.

Further reducing loads requires advanced rotor control strategies (WP5). These control strategies should be taken into account in the design of offshore support structures (WP4). The UpWind project demonstrated that individual pitching of the blades could lower fatigue loads by 20%-30%. Dual pitch as the first step towards a more continuous distributed blade control (pitching the blade in two sections) could lead to load reductions of 15%. In addition, the future smart turbine will use advanced features to perform site adaptation of its controller in order to adapt to local conditions (WP5).

Advanced control strategies are particularly relevant for large offshore arrays, where UpWind demonstrated that 20% of the power output can be lost due to wake effects between turbines.

Optimised wind farm layouts were proposed, and innovative control strategies were developed, for instance lowering the power output of the first row (thus making these wind turbines a bit more transparent for the air flow), facing the undisturbed wind, allowing for higher overall wind farm efficiency (WP8).

Control and maintenance strategies require load sensors, which were adapted and tested within UpWind. To avoid sensor failures causing too much loss of energy output, loss of sensor signals was incorporated into the control strategies (WP5) and a strategy was developed to reduce the number of sensors. The fatigue loading on individual wind turbines can be estimated from one heavily instrumented turbine in a wind farm if the relationship of fatigue loading between wind turbines inside a wind farm is known. The so-called Flight Leader Concept 4 was developed in WP7.

Those load sensors can be Bragg sensors - tested and validated within the project (WP7). UpWind demonstrated the efficiency and reliability of such sensors, and assessed the possibility of including optical fibres within the blade without damaging the structure (WP3). However, using sensors implies the rotor is only reacting to loading phenomenon. As a result of system inertia, the load will be partly absorbed.

A step further is to develop preventative load alleviation strategies by detecting and evaluating the upcoming gust or vortex before it arrives at the turbine. A nacelle-mounted LIDAR is able to do this (WP6), and can be used as an input signal for the individual blade pitching, or in distributed blade control strategies (WP5).

In recent years, UpWind has been a focal point for LIDAR development, and has considerably helped the market penetration of LIDAR technologies. Although LIDARs are still considerably more expensive than SODARs, for instance, their technical performance, and potential, is substantial. UpWind demonstrated that LIDARs are sufficiently accurate for wind applications (WP1A2). LIDARs can be used for the power curve estimation of large turbines, for control systems, for resource assessment, and for measuring the wind shear over the rotor area.

UpWind demonstrated the need to take the wind shear into account for large rotors (WP6 and WP2). The 20 MW rotor is so large that the wind inflow needs to be treated as an heterogeneous phenomenon, in contrast to more typical approaches. One point measurement, as recommended by IEC standards, is not representative anymore. A correction method was developed and demonstrated within UpWind.

The smart control strategies and high resolution modelling described require a highly accurate wind measurement, since a small deviation can have a significant impact on reliability. In the metrology domain, UpWind considerably improved knowledge on wind measurement accuracy within the MEASNET 5 community. Cup anemometers, LIDARs, SODARs and sonic anemometers (WP1A.2 and WP6) were tested, demonstrated and improved. UpWind's WP1A.2 had access to almost all existing wind measurement databases.

The advanced control strategies of smart blades using smart sensors enable loads to be lowered considerably, so lighter structures can be developed. The improved modelling capability means the design safety factors can be less conservative, also paving the way to lighter structures (WP1A1). UpWind investigated this path, developing accurate integral design tools that took into account transport, installation, and operation and maintenance (O&M).

Onshore, the transport of large blades is a particular challenge, and UpWind developed innovative blade concepts (WP1B1) enabling a component to be transported in two sections without endangering its structural safety or aerodynamic efficiency. Integral design tools were also developed to improve the reliability of the entire drive train (WP1B.2), and to investigate the possibility of developing proportionally lighter generators for large wind turbine designs. UpWind investigated 10 different generator configurations and found promising potential weight reductions for permanent magnet transversal flux generators.

The UpWind project worked on ensuring the reliability of large turbines, in particular for far offshore applications, focused on condition monitoring (WP7) and fault prediction systems. Enabling fault detection and preventative maintenance with a large potential for cutting O&M costs, the reliability of future large blades can be assessed using probabilistic blade failure simulation tools (WP3).

Reducing the loads and the nacelle weight enables the offshore substructure design to be optimised (WP4). UpWind also developed integrated wind turbine/substructure design tools and investigated optimal offshore substructure configurations according to the type of turbine, the type of soil and water depth. Future deeper water locations were investigated and innovative cost-effective designs were analysed.

Deep Water Foundations

Progress was made on deep water foundation analysis, including the development of advanced modelling techniques and enhancements of current design standards which, for example, become very important for floating designs. With the improved intelligence of wind turbines, wind farms are operated more and more as power plants, providing services to the electricity system, such as flexibility and controllability of active and reactive power, frequency and voltage, fault-ride-through or black start capabilities (WP9). Those capabilities will allow for substantially increased penetration of wind power in the grid in the near future. The future large offshore wind farms, far from shore, will be connected via HVDC VSC, forming the backbone of an integrated European offshore grid, and supporting the emergence of a single electricity market.

It will be challenging for the wind energy sector to attract and train the required number of engineers, postgraduates and PhD students to fulfil its needs. UpWind focused on training and education (WP1A3), and developed free of charge advanced training modules on wind energy, including the latest innovations in the field. This content is distributed through the REnKnow database.

A cost model was developed in order to isolate and study the dimensioning cost parameters of this upscaled wind turbine. However, as an optimal design should account for the external design constraints, an innovative design approach was developed that includes both technical and non-technical disciplines within the same framework. In this framework, manufacturing, transport, installation and O&M procedures become design parameters rather than constraints, enabling the system as a whole to be optimised at a design stage. Finally, a large potential for cost optimisation lies in the design safety levels.

A probabilistic design of structural wind turbine components can be used to design components directly, thereby ensuring the design is more uniform and economic than that obtained by traditional design using standards such as the IEC 61400 series. The challenge is to efficiently update the design standards and to promote the use of an integrated design approach. This will ensure consistency between the advanced models and strengthen their integration into wind energy technology, improve the test methods and design concepts developed in UpWind and, in turn, provide a consistent scientific background for standards and design tools.

This design approach comprises four parts: providing a reference wind turbine for ease of communication between the work packages and integration and benchmarking of their findings; developing cost models for upscaling to very large wind turbines (20 MW) - in co-operation with the upscaling work package; development and definition of an integral design method; and development of (pre)standards for the application of the integral design approach, including interfaces, data needs guidelines and proposals for a formal international standardisation process.