Research On Voltage Stability Of Grid Containing Wind Turbines

With the increased focus of environment problems and the rapid development of technology, the wind power is becoming more and more popular. With more wind farms connected to the grid and with the wind turbines capacity increasing, the influence of wind farms to the power system becomes inevitable. Static voltage stability is an important aspect of grid normal operation. Considering the fluctuation and uncertainty of wind turbine output, it's necessary to study the static voltage stability of grid containing wind turbines.In this paper, the mechanism of voltages stability is analyzed at first from both the static aspect and dynamic aspect. Some common grid static voltage stability analysis methods are also introduced. To study the voltage stability of grid containing wind turbines further, the minimum eigenvalue of Jacobi Square is used as an index of system voltage stability and the LC is used as the weakness of the nodes.The traditional way to generate correlated wind speeds in simulation of grid containing multiple wind farms is to use Cholesky decomposition to generated correlated standard normal variables. The variables are then transformed to wind speeds meeting Weibull distribution using Nataf inverse transformation. But this method is complicated and has much errors. To improve it, the improve PS algorithm instead of Cholesky decomposition method is used to generate correlated normal variables. With math derivation the effectiveness of new method is proved and with the example the superiority is also proved.Combine the method with Monte Carlo, a voltage static stability analysis method for grid containing multiple wind farms is proposed. Take the regional grid containing multiple wind farms in northeast China as an example, the characteristic of weak nodes in grid is studied and the influence of wind turbine output power factor, the wind turbines' capacity and the correlation of wind speeds at different wind farms on static voltage stability is studied. Each simulation result is also explained by corresponding physical mechanisms.