Wednesday, 20 April 2016: 4:45 PM
Ponce de Leon B (The Condado Hilton Plaza)
Widespread, disruptive power outages often occur in the wake of hurricane landfalls. The direct effect of wind on power lines, falling trees and branches, and storm surge flooding of power infrastructure can all lead to outages; the damage caused by the hurricane can also hinder efforts to restore power service. A blackout model has been jointly developed by Verisk Insurance Solutions and Hartford Steam Boiler to forecast power outages due to the wind and storm surge of hurricanes. The model can be used with real-time forecasts of a current storm or for stochastic storm catalogs. The real-time forecasts are generated using mesoscale model forecasts of wind from HWRF and forecasted surge from SLOSH; stochastic storms are modeled based on a wind footprint and surge elevation. The wind algorithm incorporates a tree cover dataset to model risk of outages caused by fallen limbs, as well as direct damage to power infrastructure from the wind itself. Additional components of the model track geographic areas that risk power disruption due to storm surge flooding of power infrastructure. This model produces percentage outages for each zip code, and an expected duration of that outage. The talk will demonstrate some aspects of how outages are forecasted, and show results from historical storms.
- Indicates paper has been withdrawn from meeting
- Submission entered in competition