Date of Award

9-2017

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Mechanical and Civil Engineering

First Advisor

Jean Paul Pinelli

Second Advisor

Ralph V. Locurcio

Third Advisor

Nezamoddin N. Kachouie

Fourth Advisor

Luis D. Otero

Abstract

Annualized hurricane-related losses to the built environment in the United States are in the billions of dollars. According to the National Oceanic and Atmospheric Administration, in 2010 39% of the United States population, or 123.3 million inhabitants, lived in a coastal county. This population is expected to increase by 8% from 2010 to 2020. The majority of these people live in buildings prone to hurricanes. The hurricanes can result in wind damage (due to high wind and/or wind-driven rain infiltration) and water damage (due to storm surge and waves or freshwater flooding) to the building and its contents, which are typically insured by separate entities. In addition to the possible damages, a hurricane can cause the occupants to flee from their homes and not return immediately, and incur time related expenses (a.k.a. additional living expenses). Catastrophe models estimate these different types of damage and expenses for entire population of buildings. Emergency managers and the insurance industry and regulators are the main users of these models. The Florida Hurricane Loss Model (FPHLM) is an example of such a catastrophe model: it is a probabilistic model which predicts the hurricane insured losses to residential buildings in the state of Florida. Originally, the FPHLM was restricted to wind and rain hazards. Recently the State of Florida commisioned an enhancement of the FPHLM to expand the scope of the model to storm surge and hurricane-related freshwater flooding. The goal of the new model is to: (1) estimate the potential damage to residential properties from both storm surge and inland flooding; (2) differentiate wind losses from storm surge/flood losses; (3) provide a state-of-the-art innovative methodology to combine the separate wind and storm surge/flood losses; and, (4) provide assistance to regulators and the insurance and reinsurance industries in the rate-making process. This dissertation describes the development of the vulnerability component of the FPHLM enhancement project. It incorporates the knowledge gained from a vast pool of studies in the fields of freshwater flooding, storm surge inundation, and tsunami research to develop the different vulnerability components for the enhancement project. In the event that a structure is subjected to damage from both wind and water hazards, the proper allocation, and subsequent combination, of damage for insurance purposes is challenging because, as in the case of the FPHLM, wind and water vulnerability are often assessed with separate models. The current literature addresses the calculation of the damages from each hurricane hazard including in some cases the combination of these damages. However, many of these methods suffer from some of the following shortcomings: (1) they focus on a specific building type only; (2) they relate damages to a specific hazard intensity without consideration of all acting forces (e.g., inundation depth without taking into account wave action and water velocity from surge-induced damages); (3) they base their results on limited damage data (e.g., most empirical vulnerability models); (4) they limit damages of specific components to a particular hazard (e.g., limiting structural damages to wind only); or, (5) they make assumptions that are not realistic (e.g., the independence of wind-induced and water-induced damages). To address these shortcomings, as part of the enhancement project, this dissertation presents a method to characterize and quantify storm surge damage states for residential buildings. These damage states are then used in the generation of fragility and vulnerability curves. Next, the dissertation presents a new methodology for calculating the water-induced damage to the content and the additional living expenses due to storm surge or fresh water inundation. Finally, it describes a methodology for the assessment of multi-hazard vulnerability in which the damage from each hazard (wind and rain, storm surge, and freshwater inundation) is calculated separately. An engineering approach in conjunction with hazard information (intensity, timing) and statistical methods is the basis for the calculation of the combined damage and the assignment of damage causation from each hazard. This dissertation also includes an uncertainty analysis that evaluates the reliability of the outputs of the different components of the FPHLM and the variability of the inputs based on the model assumptions. The FPHLM development team has access to the exposure and claim data of major private insurance companies and the National Flood Insurance Program for all hurricanes that have impacted the state of Florida since 1975. It also has access to Florida tax appraiser databases that define the existing building inventory. These datasets were used in the development of the surge and flood content damage model as well as the validation and calibration of other components in the FPHLM. The use of these datasets required extensive processing and interpretation. This dissertation discusses the challenges encountered and the solutions proposed in this process.

Comments

Copyright held by author

Download

Share

COinS