Extreme Event Modeling

Catastrophic situations, such Superstorm Sandy, Hurricane Katrina, and the 2003 Northeast blackout, pose an enormous threat to the nation's electric grid and the socio-economic systems that depend on reliable delivery of power.   Utilities are unable to adequately plan and prepare for such events, despite the economic costs and social hardships they cause.

One primary bottleneck that impedes progress for such planning and preparation is the ability to model extreme events.  In this project, we address two key aspects of extreme events:  Cascading and N-k.   We use a novel combination of existing capability, high-performance computing, and new capability to:

• improve the computational efficiency of models and simulations of cascades
• address modeling inadequacies
• incorporate probabilistic approaches to deliver significant advances in extreme event modeling.

We are delivering a prototype tool that simulates cascades up to 500 times faster than existing approaches.  Current tools are 10-100 times slower than real time—leaving planners with insufficient time to analyze and prepare for high-consequence events.  This tool includes capabilities for probabilistic N-k, and executes from existing trusted commercial power software.