NEW DIRECTIONS IN POST-EARTHQUAKE FIRE HAZARD ANALYSIS WITH APPLICATIONS TO MIDWESTERN UNITED STATES
MetadataShow full item record
Post-earthquake fire ignition (PEFI) can lead to severe structural damage following an earthquake. Estimating the risk of such ignitions in buildings and identifying methods to abate it are essential steps in an overall effort to mitigate the impact of post-earthquake fires in urban areas. While several models have been developed for areas with available historical PEFI data, such as the Western United States, no such models have been developed for areas with little or no data specific for post-earthquake fires. Examples of such areas are seismic regions in the Midwestern United States. The lack of PEFI data for these areas is due to the fact that at the time of several significant earthquakes in the early nineteenth century, most earthquake-stricken communities where rural. With the growth of urban areas in the region, a need exists for a methodology that can be effectively used in estimating PEFI risks when no or little historical data is available. In this research, it was found that models for PEFI risk estimation may indeed be developed using available data on ignitions for normal conditions as a basis and then using some type of a modification factor to account for the significance of future earthquakes. This modification factor depends on the characteristics of the region in terms of seismic activities and the type and distribution of buildings and their potential in promoting ignitions. The term “normal condition ignition” (NCI) refers to an ignition that occurs due to everyday activities and routine operations in a building. In a residential building, such activities include, for example, operating heating units and burners, cooking, and mechanical malfunction of appliances. In this research, it was found that four factors specifically affect PEFI risk and can be used to develop models for risk estimation. These are (1) spatial characteristics such as geographic concentration of particular building types; (2) ignitability characteristics, such as the sources of ignitions in a particular building type, (3) earthquake characteristics (such as the peak ground acceleration); and (4) temporal characteristics, such as the time of the earthquake and seasonality. Accordingly, models for estimating the risk of post-earthquake fire ignition occurrence are developed. These models are tested, and the model parameters calibrated, using information in areas for which both the NCI and PEFI data are available (such as in the Western United States). To illustrate the applicability of the models developed and proposed in this study, St. Louis City is considered. This constitutes a major urban area vulnerable to potential future seismic activities because of proximity to the New Madrid Fault Zone. Using the NCI data for this area, PEFI risk values are estimated based on probable future seismic activities in the region. The results are presented in terms of estimated annual risk of post-earthquake fires for the area specifically for residential buildings (such as single or multifamily dwellings). The study further discusses the significance of PEFI models, their limitations and also provides suggestions for the future continuation of the research.