Model Reduction of Induction Machines and Improvement in Power Quality of Doubly Fed Induction Generator Based Wind Turbines

Abstract

Wind energy is recognized as an unlimited source of clean and non-polluting renewable energy. To reduce the dependence on fossil fuels for electric power generation, developing countries like Pakistan have opted for renewable energy solutions, especially wind energy. However, wind energy conversion systems are challenged with power quality issues. The advancement in technology has made the operator and consumer power quality conscious. These power quality issues are being addressed by introducing standards and grid codes. Initially the research presents a computer simulation-based performance comparison under different wind speeds and fault conditions for Doubly Fed Induction Generator, Permanent Magnet Synchronous Generator and Squirrel Cage Induction Generator. The measures used for performance comparison are Total Harmonic Distortion, generated power & voltages, and rotor speed. As compared to Squirrel Cage Induction Generator and Permanent Magnet Synchronous Generator, Doubly Fed Induction Generator is found to have the least contribution to THD at the grid, achieves the set value of terminal voltage in minimal time, and provides stable output power in transient period. A Merit function developed on the basis of statistical analysis of the observed THD data also supports the simulation results. The simulation parameters match the actual parameters at Wind Power Project, Jhimpir, Pakistan. Wind turbines are arranged into smaller groups called clusters in a wind farm to economize land utilization and the length of medium voltage collection power cable. The simulations are extended to investigate the effect of cluster size on power quality. Since DFIG was found better than the other two generators, a computer simulation model of a DFIG based wind farm is used for the investigations of further mitigation of THD. Sapphire Wind Power Project at Jhimpir, Pakistan, consists of thirty-three Doubly Fed Induction Generator (DFIG) based wind turbines. Cluster size varying from 4-wind turbines to 9-wind turbines for three different WTGs of output powers of 1.5 MW, 1.62 MW and 2.0 MW is considered. Each configuration is simulated for a normal (i.e., no-fault) condition at the grid and a three-phase-to-ground fault (i.e., worst) condition. It is observed that increasing the size of the cluster results in decreasing total harmonic distortion. However, the optimum length of the cable for medium voltage collection limits the maximum cluster size. Simulations further show that wind turbines with larger power decreases total harmonic distortion. The simulation results are supported by the evaluation of the Merit function established on the basis of statistical analysis of the observed data. The v simulation results are insightful and provide guidelines for the improvement of Power Quality of a wind farm through mitigating the THD at design level by the appropriate selection of Wind Turbine Generator, Cluster Size in a wind farm and Power output of Wind Turbine Generator. SCIG needs a high starting current and produces a low starting torque. Therefore, SCIG uses a double cage rotor, also called Double Squirrel Cage Induction Generator, to address the said problems. DCIG has higher-order complexity than the single rotor SCIG, DFIG and PMSG especially when it forms a cluster in a wind farm. Hence, the computer simulation-based study, investigations, and research demands a perfect depiction of Double Cage Induction Machine that can address their significant issues, especially of model order complexity, computational efficiency, controllability, observability and issues concerning integrating wind energy conversion systems to the grid. The state-space representations grant a convenient, compact, and elegant way to examine the physical systems with facts readily available for stability, controllability, and observability analysis. Hence, the research is stretched to present the model order reduction of a stable DCIM based variable-speed wind turbine model. It is performed with the aid of the proposed stability preserving balanced realization algorithm based on discrete frequency weights and limited frequency-interval. The proposed approach not only ensures the stability of the reduced-order model but also provides low approximation error as compared with other existing approaches and also provides an easily calculable a priori error bound formula. The proposed work produces steady and precise outcomes in contrast to conventional reduction methods, which shows the efficacy of the proposed algorithm