Power smoothing in large wind farms using optimal control of rotating kinetic energy reserves

In large wind farms, self-induced turbulence levels significantly increase the variability of generated power in a range of time scales from a few seconds to several minutes. In the current study, we investigate the potential for reducing this type of variability by dynamically controlling the rotating kinetic energy reserves that are present in the farm's wind turbines. To this end, we reduce the burden of frequency regulation on remaining conventional units when they are displaced in favor of wind turbines. We focus on the development of a theoretical benchmark framework in which we explore the trade-off between high energy extraction and low variability using optimal coordinated control of multiple turbines subject to a turbulent wind field. This wind field is obtained from a large-eddy simulation of a fully developed wind farm boundary layer. The controls that are optimized are the electric torque and the pitch angles of the individual turbines as function of time so that turbines are accelerated or decelerated to optimally extract or store energy in the turbines' rotating inertia. Results are presented in terms of Pareto fronts (i.e., curves with optimal trade-offs), and we find that power variations can be significantly reduced with limited loss of extracted energy. For a one-turbine case, such an optimal control leads to large potential reductions of variability but mainly for time scales below 10s if we limit power losses to a few percent. Variability over longer time scales (10-100s) is reduced considerably more for coordinated control. For instance, restricting the energy-loss incurred with smoothing to 1%, and looking at time scales of 50s, we manage to reduce variability with a factor of 6 for a coordinated case with 24 turbines, compared with a factor of 1.4 for an uncoordinated case. Copyright (c) 2014 John Wiley & Sons, Ltd.