Modeling vulnerability of residential buildings to multiple hazards

The majority of Florida's 19 million residents live at or near a hurricane- or tornado-prone area, and their homes are very susceptible to these extreme wind climatology. Nevertheless, homeowners have very little or no information on the risk of their houses until hurricanes or tornadoes come through. In order to provide public demonstration of a community's vulnerability to these hazards, reliable and robust damage and loss projection models are needed. Further, insurance companies also require a means of assessing hurricane and tornado risk and generating loss estimates that can be used within the actuarial rate-setting process. It is also apparent that public policy officials require methods to predict the potential economic loss from natural hazards in order to aid in identifying high-risk region, implementing necessary damage mitigation measures and preparing their populations and response for future natural hazard events. This research develops new approaches to assess the vulnerability of residential buildings to hurricanes and tornadoes.;Coastal constructions often sustain damage from both wind and surge during hurricane landfalls, and while it is known these forces can affect residential construction, less is known about the interaction of wind forces and storm surge events. Current models do not provide methods for predicting the damage from the joint probability of a wind and water event. The Florida Public Hurricane Loss Model (FPHLM) sponsored by the Florida Office of Insurance Regulation already includes existing modules to predict wind damage and as part of a project to enhance the features of that model, this research was conducted to develop an engineering-based storm surge and inland flooding vulnerability component to its existing wind-only model.;To achieve this goal, this study develops vulnerability functions of three types of residential buildings under four hydraulic conditions: inland flooding, storm surge with minor waves, moderate waves and severe waves. The study's subject houses include slab on grade personal residences, elevated structures and manufactured homes. For personal residences, fragility curves are developed using a semi-engineering methodology and a translation method is employed to convert them into vulnerability functions. Additionally, the vulnerability curves of manufacture houses and structures elevated on piles are generated with a combination of previous claimed data, engineering judgement and the forgoing semi-engineering approach. The set of vulnerability curves generated in this study could be directly combined with the FPHLM wind model for the estimation of economic losses from hurricanes.;An engineering-based tornado damage assessment (ETDA) framework for a typical light-frame wood residential structure is presented in Chapter 4. It includes a translating tornado vortex model, an analytical tornado-induced load calculation approach, a probabilistic methodology for wind-borne debris impact, and a time-variant internal pressure model. The framework enables users to determine structural damage to a building through successive time steps as a tornado translates past the building. The ETDA model is illustrated using four houses damaged in the 2011 Joplin, MO tornado. The damage to the four houses predicted by the ETDA model agree well with the observed damages obtained from a ground survey, with overall damage ratios differing by 15%. To further examine its rationality, the ETDA framework are applied to a portfolio of single-family residential dwellings (778 homes) damaged in the 2011 Joplin, MO tornado. The degree of damage (DOD) of each house is estimated matching the outputs of the ETDA model to the revised descriptions in the Enhanced Fijta scale and compared against those based upon the observations in a damage survey. The comparison shows reasonably good agreement and 78% of the homes show no more than +/-2 DOD rating differences.