Abstract

Modern wind systems generate electrical energy running at variable rotational speed. Operating as energy generation plants they present a stochastic behavior dependent on random wind speed. So operation performance becomes an issue when it is approached integration into the grid either in terms of quality of delivered power or in terms of generated power available to the Transmission System Operator. The communication deals with real performance of two types of wind systems integrated into a wind farm. Both the systems are modern ones, running at variable rotational speed, being one based on synchronous electrical generator, another one on asynchronous electrical generator. Approaching a wind system as a production unity whose output is dependent on a stochastic input the authors analyze behavior of the wind systems with the method "Data Envelopment Analysis". The performed analysis allows at concluding on optimization methods not only driven by theoretical analysis of the wind system but by real operation of the systems. According to global behavior in terms of energy generation the authors discuss some of the reasons for the results found. They conclude pointing out needs of integrating control actions for the wind turbine and the electrical generator based on getting knowledge on the system through data preprocessing and pos-processing allowing at improving global operation. ©2009 IEEE.