Hurricanes have caused severe damage to the electric power system throughoutthe Gulf coast region of the U.S., and electric power is critical to post-hurricane disasterresponse as well as to long-term recovery for impacted areas. Managing hurricane risksand properly preparing for post-storm recovery efforts requires rigorous methods forestimating the number and location of power outages, customers without power, anddamage to power distribution systems. This dissertation presents a statistical poweroutage prediction model, a statistical model for predicting the number of customerswithout power, statistical damage estimation models, and a physical damage estimationmodel for the gulf coast region of the U.S. The statistical models use negative binomialgeneralized additive regression models as well as negative binomial generalized linearregression models for estimating the number of power outages, customers without power,damaged poles and damaged transformers in each area of a utility company?s servicearea. The statistical models developed based on transformed data replace hurricaneindicator variables, dummy variables, with physically measurable variables, enablingfuture predictions to be based on only well-understood characteristics of hurricanes. Thephysical damage estimation model provides reliable predictions of the number ofdamaged poles for future hurricanes by integrating fragility curves based on structural reliability analysis with observed data through a Bayesian approach. The models weredeveloped using data about power outages during nine hurricanes in three states servedby a large, investor-owned utility company in the Gulf Coast region.
展开▼
