| The blackout is the most serious form in power system accidents. Since the new century, researchers have found that the scale of blackouts around world obeys a power law distribution rather than the classical exponential of theoretical assumptions. Studies have shown that the power system as a complex system with self-organized criticality. Blackout as the dynamic behavior of the power system, its scale has power-law distribution and long-range autocorrelation. Under the impetus of external disturbances and load growth, the power system will continue approaching self-organized critical state. In this critical state, the external disturbances could lead line trip, the loads flowing could trigger a cascade overload and invisible fault protection, expand the scope of the accident and result in a power-law distribution characteristics in blackouts. Blackouts often happen with a large number of failures. Research shows that the power grid failures and blackouts as power-law distribution properties and has long-term autocorrelation. But there are no papers which power-law distribution grid failure characteristics of the blackout has what role. Likewise, the grid failures behave the mechanism of the long-range correlation, but also a lack of relevant research. This paper will research for such problems.In order to investigate effects of external disturbance caused a power law characteristics of the blackout, the article get the recorded data from the North American and China blackout data to find the evidence of external disturbance independent leading to blackouts power law characteristics. First, from differences of power-law distribution and exponential distribution, to analysis 29 extreme blackout records of the North American blackout that caused a power law characteristic thick tail portion and found that 10 cases occurred in adverse weather conditions and is just in the distribution network. Because of the distribution network for the ring network design but open-loop operation, the probability of cascade overload is much lower than the grid, which may be caused by part of the blackout extreme external disturbance rather than cascading failure. Secondly, this paper also analysis the domestic 9 network which in characteristics with a power-law distribution. Jiangxi data showed that only half of the load under normal circumstances during the ice storm, there is not cascade overload. In addition, this province got line trip 146 times during the ice storm, but only once to determine the relay malfunction, there are no obvious cascade fault. The ice storm blackout is Jiangxi province’s largest power outage, which proved that the outside disturbance can cause blackouts characteristic power law in the absence of cascading failure situation as well.To verify the peak failure from bad weather caused the long-term correlation of grid failure, the article also analyzed the data of grid failure in Changsha and Nanchang. The fault data on a daily basis the number of failures, to exclude descending extreme maximum daily failure data step by step, and using the fluctuation method to calculate the trends for each data series of Hurst exponent, Hurst values have found which is decrease with excluding maximum fault day, when maximum of daily failure is 10 or 9, Hurst exponent value even less than 0.5. This means that the long-term power failures autocorrelation is mainly caused by peak of fault which is from the external disturbance such as bad weather. |
PDF Full Text Request