Study On Reliability Evaluation Models And Algorithms Of VSC-HVDC Systems For Grid Connection Of Wind Farms

With the increasingly serious energy shortage and environmental problems, countries all over the world are competing to increase the development and utilization of renewable energy. Wind energy has become one of the most promising renewable energy with the advantages of rich resources, clean and no pollution. Whereas wind energy also has the characteristics of intermittence and volatility, and with the increasing scale of wind farms connected to the power grid, the whole power system’s reliability and stability is affected more obviously when adopting AC grid connection. Research shows that many problems including the voltage fluctuation caused by the fluctuation of power output of wind farms can be alleviated and overcame to some extent by VSC-HVDC, especially multi-terminal VSC-HVDC, with its flexible and controllable characteristic, which makes it become the future development direction of grid connection of wind farms. However, the current study on reliability evaluation model and algorithm of VSC-HVDC systems for grid connection of wind farms is relatively deficient, so this thesis makes a special study on this problem. The main study contents of this thesis are as follows:In the reliability evaluation of wind farms, frequency characteristic of wind speed fluctuation is not included in the traditional multi-state model of wind turbines, which makes it unable to calculate the frequency index of system. Although sequential Monte Carlo method can solve this problem, but it usually needs a longer simulation time, which makes it not conducive for the actual engineering application. Therefore, a multi-state Markov model of wind farms is established based on analytical method. Firstly, sequential power output of a single wind turbine is determined based on the ARMA model of wind speed and wind power conversion characteristic; Secondly, the power output of a single wind turbine is divided into some states by using K- means clustering method, according to the state transferring relationship between wind speed and wind turbines, a Markov model of a single wind turbine is established; Finally, by clustering the power output of wind farms which is combined with n wind turbines, based on the boundary wall principle to determine the transfer rate between states, a seven-state Markov model of wind farms is established. Case study shows that the proposed algorithm can quickly calculate the probability and frequency index of wind farms in different states.Currently, the study on the reliability model of VSC-HVDC systems is mainly based on the existing reliability model and method of traditional HVDC systems, ignoring the reactive power compensation capacity of VSC-HVDC systems working as STATCOM and the impact of intermittence of wind power output. Therefore, on the basis of intensive study on the main components and failure consequences, the Markov models of components including transformers, converters and so on are established. According to the effect of component failures on the system state transition, a unified Markov model of VSC-HVDC systems is established considering STATCOM state and intermittence of wind power output. Case study shows that, the evaluation results can more accurately reflect characteristics of reactive power compensation of VSC-HVDC systems considering STATCOM state. The reliability index has been greatly improved considering intermittence of wind power output.The operation mode of multi-terminal VSC-HVDC systems for grid connection is relatively complex, so reliability modeling directly based on analytical method is difficult and unable to consider the sequence correlation relationship between wind farm power output and component failures of VSC-HVDC systems. To overcome these problems, a hybrid reliability evaluation method is proposed for multi-terminal VSC-HVDC systems combining Markov method and Monte Carlo method. Firstly, the grid connection system will be divided into different subsystems according to topology characteristics of multi-terminal VSC-HVDC systems and functions of each component, including subsystems of wind farm, rectifier terminal, inverter terminal and transmission, and Markov model of each subsystem is established. Then the reliability index of systems can be calculated by sequential Monte Carlo method. In addition, a reliability evaluation index framework is established in order to describe the impact of intermittence of wind power output and the reactive power compensation capacity of VSC-HVDC systems. Case study shows that the proposed hybrid method can effectively reduce the complexity when reliability evaluating, and reliability evaluation results can consider sequence correlation relationship between wind farm power output and component failures of VSC-HVDC systems.