An elemental study of optimal wind power plant control
Karl Merz, John Olav Tande
SINTEF Energy Research, Trondheim, Norway
To control a wind power plant optimally requires knowledge of how the control actions impact the wind turbines, atmosphere, and electric grid. Any control strategy must deal with performance tradeoffs between the control objectives of load reduction, maximum energy production, and controllable power for electric grid services. We explore the dynamics of a wind power plant, using an integrated multidisciplinary model, in order to study these tradeoffs and examine what the "optimal" control of the plant may be.
Taking an incremental approach, we move from a model of a single 10 MW offshore wind turbine, to a "wind power plant" consisting of two turbines. The wind turbines are outfitted with a control system that provides all the functions expected of a modern turbine operating as part of a large plant: active damping of the driveshaft and tower modes, load rejecting functions including individual blade pitch, and operator power command tracking. A simple atmospheric model, building on recent results in the field of wind turbine wake dynamics, couples the turbines through the atmosphere. The turbines are also coupled by a model of an electric collector grid, and wind power plant control system.
The two-turbine model is used to study the influence of the control strategy on the modal dynamics and performance of the system. We find that the choice of plant-level control strategy has a significant influence on the dynamic properties of the system. We also expect that the majority of solutions tailored towards load reduction or grid services will end up reducing energy production to the extent that it does more harm than good to profitability. Opportunities for providing load reduction and grid services, while maintaining energy production, will be identified.
Optimal operation of wind power plants will result in greater profitability. Production may be marginally increased, through adapting the turbine control to flow and wake conditions. At the same time, the costs associated with scheduled and unscheduled maintenance can be reduced, through tailoring the operating strategy to the turbines' condition. Holistic dynamic models of wind power plants, such as the one described here, are the basis for state estimation, which in turn is applied in condition monitoring and real-time optimal control.
Delegates will gain an understanding of methods which can be used to construct dynamic models of large wind power plants, suited for condition monitoring and control applications. The model is of a type which is simple enough to be readily understood, via modal dynamics, yet it includes a full description of each wind turbine: aeroelastic, electrical, and control. The influence of an optimal control strategy is demonstrated through case studies.