Control algorithms for effective operation of variable-speed wind turbines

Abstract

This report describes a computer code, called ASYM and provides results from its application in simulating the control of the 34-m Test Bed vertical-axis wind turbine (VAWT) in Bushland, Texas. The code synthesizes dynamic wind speeds on a second-by-second basis in the time domain. The wind speeds conform to a predetermined spectral content governed by the hourly average wind speed that prevails at each hour of the simulation. The hourly average values are selected in a probabilistic sense through the application of Markov chains, but their cumulative frequency of occurrence conforms to a Rayleigh distribution that is governed by the mean annual wind speed of the site selected. The simulated wind speeds then drive a series of control algorithms that enable the code to predict key operational parameters such as number of annual starts and stops, annual energy production, and annual fatigue damage at a critically stressed joint on the wind turbine. This report also presents results from the application of ASYM that pertain to low wind speed cut-in and cut-out conditions and controlled operation near critical speed ranges that excite structural vibrations that can lead to accelerated fatigue damage.