Comparative Analysis of Blackout Data from North America Power Grids Using the Highly Optimized Tolerance Model

Xiangning Lin,Zhiqian Bo

DOI: https://doi.org/10.1109/pes.2008.4596203

2008-01-01

Abstract:The Highly Optimized Tolerance (H.O.T.) model which aims to describe the statistics of robust complex systems in uncertain environments, is used to analyze the electric blackouts occurred on North America Power Grids. The cumulative probability of event size is taken into account as the index observing the law that the blackout follows. As the assumption of the existing work, the resources are optimally distributed with respect to the megawatts lost if the standard H.O.T model is well fit the statistic data. Therefore, the data from the Western power grid (western system coordinating council, W.S.C.C.), together with those from northeast power grid(Northeast Power Coordinating Council, N.P.C.C), ranging from 1984 to 1999 of all power outages occurring on the grids are used for comparison. These data are measured in the size of megawatts lost In applying the model to the power grid data, we observe that the model agrees closely with the W.S.C.C. data but not exactly the N.P.C.C data. One of the possible explanations is that that the optimal resource distribution of power grid is valid for the W. S. C. C but not the N.P.C.C. However, another possibility is that the existing standard H.O.T model is still not a generic model in terms of power system application. In this sense, a power-system-objected model is worth studying for the purpose of direct instruction using H.O.T. theory as the tool of power system blackout mechanism analysis and countermeasure.