DRC-Disaster Research Center

Active Studies










Modeling the Ignition and Spread of Post-Earthquake Fires

Principal Investigators: Rachel Davidson (Co-PI)
Graduate Research Assistants: Sizheng Li
Funding Agencies: Unfunded

Study Description

For fifty years, post-earthquake fire spread models used the basic empirical, macro approach developed by Hamada in 1951. Recently, however, a few physics-based models have emerged in which different modes of fire spread (e.g., radiation, branding) are modeled separately, more realistically, and in more detail. In this project, we have developed a new physics-based model that simulates the spread of post-earthquake urban fires. Adapting and integrating compartment fire models, this new model explicitly represents: (1) the evolution of fire within a room or roof; (2) room-to-room spread within a building through doorways, by burning through walls and ceilings, and by leapfrogging between windows; and (3) building-to-building spread by window flame impingement; radiation from window flames, room gas, and roof flames; and branding. Using a case study for Los Angeles, this research illustrates how the new model can be used to estimate how much, how fast, and where fires spread, and also to provide insight into how fires spread and factors that influence fire spread.

In another effort as part of this project, we have developed a new approach to statistical modeling of post-earthquake fire ignitions and to data collection for such modeling, and applied it to late 20th century California. Generalized linear and generalized linear mixed models (GLMs and GLMMs) are used for this application for the first time. The approach recognizes that ignition counts are discrete, examines many possible covariates, and uses a small unit of study to ensure homogeneity in variable values for each area unit. Two datasets were developed to explore the effect of missing ignition data, each with a different assumption about the missing data. For one dataset, the recommended model includes instrumental intensity; percentage of land area that is commercial, industrial, or transportation; total building area; percentage of building area that is unreinforced masonry; and people per sq. km. The other includes the same, except area of high-intensity residential development replaces total building area, and median year built over all housing units is also included. The models should be useful in estimating the number and locations of post-earthquake ignitions in future earthquakes.

Study Related Publications

Lee, S., and Davidson, R. Physics-based simulation model of post-earthquake fire spread, in review.

Lee, S., and Davidson, R. Application of a physics-based simulation model to examine post-earthquake fire spread, in review.

Lee, S., Davidson, R., and Lembo, A. Application of advanced GIS algorithms to support post-earthquake fire spread modeling, in review.

Davidson, R. Modeling Post-earthquake Fire Ignitions using Generalized Linear (Mixed) Models. Journal of Infrastructure Systems, in press.

Davidson, R. 2008. Generalized Linear (Mixed) Models of Post-earthquake Ignitions. MCEER technical report, MCEER-09-0004, Buffalo, NY, in press.

Lee, S., Davidson, R., Scawthorn, C., and Ohnishi, N. 2008. Fire following earthquake– Review of the state-of-the-art modeling. Earthquake Spectra, 24(4), 1-35.

Study Related Presentations

Davidson, R. Generalized linear (mixed) models of post-earthquake fire ignitions. Proc., 14th World Conference on Earthquake Engineering, October 12-17, 2008, Beijing.

Lee, S., and Davidson, R. Modeling different modes of post-earthquake fire spread. Proc., 14th World Conference for Earthquake Engineering, October 12-17, 2008, Beijing, Paper no. 13-0014.

Lee, S., and Davidson, R. Simulation-based model of fire following earthquake. Proc., 8th U.S. National Conf. on Earthquake Engineering in San Francisco, CA, April 18 - 22, 2006.