Reliability Evaluation Of Distribution System With Distributed Generators

Distributed generator (DG), with its flexible power generation, environment-friendly advantages has been increasingly applied to the distribution system, which has brought some influence to the traditional distribution network reliability evaluation. Compared with the traditional distributed power supply, the output power of DG has certain randomness. In addition, the traditional evaluation model is a radial network, in which the feeders are usually supplied by single power source. When there's a failure of any component of the feeder, the load points collected with the feeder would be out of energy. When DG is connected to the distribution system, the radial network will turn into a multi-source network, and the evaluation model and algorithm are different with the traditional one when DG is operated in island mode.The reliability evaluation model of DG is established on the basis of fully studied the characteristics of DG in this paper. It takes micro-gas turbine as an example to establish a non-intermittent distributed power reliability evaluation model, describing as a two-state model and the wind turbine generator (WTG) as an example to establish the reliability of distributed power intermittent assessment model. Fully considering the property of intermittence and randomness of wind power and by use of statistic information of wind speed in one year and according to power characteristic of WTG, a method to obtain output power of WTG and then to achieve multi-state model of WTG is proposed. In this paper, two types of reliability evaluation algorithms are improved to evaluate the reliability indexes of distribution system with DG, the minimal cut set algorithm and the sequential Monte Carlo simulation, these two algorithms each with distinct characteristics, different advantages, can be used to evaluate the reliability of distribution system with DG accurately.Finally, a typical system is applied to testify the efficiency and the accuracy of the improved models and algorithms. The results also show that DG can improve the distribution network reliability.