Print

Balasubramaniam, Karthikeyan

Electrical Engineering

2010

MS

Analysis of blackout data from the North American Electric Reliability Council (NERC) shows power-law distributions for various measures of blackout sizes, such as the amount of power lost and the number of customers unserved. Several recent papers report on simulation models that similarly produce power-law in blackout sizes. Based on these observations it has been suggested that power grids show criticality. The implication of power-laws in the probability distribution of blackout sizes is that there is a substantial increase in the risk associated with large blackouts when compared to Gaussian distributed blackout sizes. The goal of this thesis is to build a non-linear dynamical power system model, the Dynamic Cascading Failure Blackout Model (DCFBM) that captures electrical, mechanical and discrete dynamics of power grids and to use the model to further test the hypothesis that power grids show evidence of criticality. Also, several recent papers provide evidence that systems close to critical transition show 'critical slowing down'. I use DCFBM, to test 'critical slowing down' in the IEEE 9-bus test case and find results that suggest critical slowing down.