Improvements in the modeling of wind turbines in power system studies

Abstract

Several methods to improve the current wind turbine modeling techniques for power system studies are discussed in this paper. Some of the modeling improvements which are discussed are dynamic initialization of models, differences between wind turbine models developed in electromagnetic transient program and bulk power system simulation software, and techniques to avoid numerical instability with minimum errors. Algorithms to provide automatic initial conditions of all variables for any given load flow condition, wind speeds, and wind farm control strategies are developed in this paper. This approach is an improvement to the traditional method of providing constant initial values for given operating conditions. Several physical components such as harmonic filters, cable impedances, and complex control algorithms are either simplified or excluded in the power system models. The fault response of the power system model under extreme conditions leads to a numerically unstable condition. The instability might be a result of simplification discussed above and does not occur in the real turbine. Additional modeling techniques that can be used to achieve numerical stability, dynamic and fault response close to the actual wind turbine are discussed. These techniques are used to compensate for the simplifications made in power system models. Finally, the issue of nuisance tripping faced very frequently by the Independent System Operators, while working on the studies involving wind power plants, is addressed. Various methods to reduce the deviations and prevent nuisance tripping are implemented in the paper.