Initial Evidence for Self-Organized Criticality in Electric Power System Blackouts

We examine correlations in a time series of electric power system blackout sizes using scaled window variance analysis and R/S statistics. The data shows some evidence of long time correlations and has Hurst exponent near 0.7. Large blackouts tend to correlate with further large blackouts after a long time interval. Similar effects are also observed in many other complex systems exhibiting self-organized criticality. We discuss this initial evidence and possible explanations for selforganized criticality in power systems blackouts. Selforganized criticality, if fully confirmed in power systems, would suggest new approaches to understanding and possibly controlling blackouts.

• Type: Conference
• Year: 2000
• Where: HICSS
• Authors: Benjamin A. Carreras, David E. Newman, Ian Dobson, A. B. Poole
• Keywords: Biometrics, HICSS 2000, Large Blackouts, Long Time, Selforganized Criticality