A hurricane simulation method for Florida utility damage and risk assessment

The 2004-2005 hurricane season caused severe infrastructure damage and economical loss to power systems. A good estimate of hurricane frequency, intensity, and duration is of great importance to emergency planners of utilities in vulnerable coastal areas. However, considerable uncertainty in hurricane characteristics makes the modeling a complicated task. This paper presents a probabilistic model that can simulate a hurricane year for Florida. This model determines the number of hurricanes landed in Florida in a year; and each simulated hurricane is probabilistically assigned landfall information such as location, approach angle, translation velocity, maximum wind speed, and radius to maximum wind. As the hurricane moves across Florida, its inland features including wind speed decay rate, central pressure filling rate, wind field profile, and duration are further simulated. Through Monte Carlo simulation, this hurricane model is able to generate a large number of random hurricane years so as to estimate the long-term risk levels arising from hurricanes and assist associated planning. The validation of the simulated wind map against the actual ASCE 7 wind map demonstrates that this hurricane model can well reproduce the actual pattern of overall hurricane activities in Florida.