Intelligent Networking – Wind Manager
Answers in the Wind
Offshore wind parks are playing an increasingly important role in power generation. Siemens has developed a concept that could enable them to operate much more efficiently.
The world market for wind power currently amounts to 6 billion € and is growing at about 13 % annually. In Germany, wind energy, which has a power output of almost 17,000 MW, accounts for 4 % of the country’s electric energy production, and that figure is rising. By 2030, plans call for the installation of offshore wind parks on the North Sea and the Baltic Sea that will generate some 20,000 to 25,000 MW. Located far from towns, they offer a range of advantages.
A measuring station installed on the North Sea in 2003 has shown that prevailing winds have speeds of more than 4 m/s—the minimum wind speed the rotors require—95 % of the time. For three months of the year, the winds are so strong that the facility reaches its maximum rated output; inland wind parks enjoy similar conditions for only six weeks a year.
In 2004, Siemens Power Generation (PG) purchased Danish wind turbine manufacturer Bonus Energy A/S, thereby strengthening its stake in the wind energy market. Bonus Energy has installed more than 5,000 turbines in over 20 countries with a combined output of more than 3,000 MW, including the world’s largest offshore wind park 10 km south of Nysted. The total output of all 72 wind turbines in this installation is around 166 MW, enough power to cover the needs of 145,000 households.
A study carried out by the German Energy Agency in 2005 shows how the energy generated by the offshore wind park could be effectively fed into the grid. For example, the high-voltage grid could be expanded by 850 km, or storage stations could be built at sea with a shared cable connecting them with the coast. Servicing the windmills poses a particular challenge, because it’s difficult, expensive and sometimes impossible to make repairs on the high seas. As a result, offshore wind turbines have longer down times than those on land and are thus less efficient. It’s therefore particularly important to identify malfunctions in offshore parks at an early stage—ideally, before a breakdown occurs.
A wind park in the sea near Nysted, Denmark. In the future, a network of sensors could further increase energy yields and lower operating costs
Siemens Corporate Technology (CT) has developed a concept for a smart wind park equipped with a sensor network that could monitor itself autonomously. Here, wirelessly networked sensors located on the masts and rotor blades or inside the turbines would measure factors such as wind force, wind direction and vibrations. A computer on each mast would evaluate the data. These computers would diagnose malfunctions and actively intervene via actuators. On the basis of wind force readings, the system could predict sudden gusts of wind and make the necessary adjustments to the rotor blades, thus reducing the strain on the turbine.
Because the computers on the masts would communicate with one another, individual sensor data could be used by the entire park. The turbines in the first row to be hit by the wind would transmit their information to the windmills in the rows behind so that the latter could make anticipatory adjustments for optimized operation. Such a system would help to optimize wind park management in other ways. "With this intelligent control system, the windmills in the first row could be adjusted to provide better flow properties for the turbines behind them," says Prof. Martin Greiner from CT. "As a result, the power output of the wind park as a whole would be higher than it would be if each windmill optimized only its own output. That makes it possible to manage the amount of power being fed into the power grid at which times."
This concept resulted from Greiner’s work on self-organizing communication networks (see Smart Grains of Sand). Greiner, a physicist, based his research on his extensive university experience with the random modeling of turbulent flows. "The research and development work, implementation and trials would take about three years," says Greiner. "The concept could thus be implemented in 2008."
Sylvia Trage