This paper investigates one class of roadway accidents, motor vehicle fires. It provides an overview of the different data sources at both the state and the national levels. Two different types of motor vehicle fires are investigated (fatal crash fires and non-crash fires). Based on the analysis of two years of FARS data, fire occurrence in fatal crashes was found to be approximately 28 fires for every 1,000 vehicles involved. Based on fire occurrence in fatal crashes, the analysis indicated that Volvos are considered to be one of the safest motor vehicles. The paper provides four simple linear regression models for predicting motor vehicle fire rates in fatal crashes. The first model indicates that 86 percent of variation in motor vehicle crash fire rates can be predicted using vehicle age. The second and the third models show that 69 and/or 60 percents of motor vehicle crash fire rates can be predicted using the travel speed by which the vehicle was travelling before the occurrence of the accident. The fourth model indicates that approximately 22 percent of the variation in fatal crash fire rate can be explained by the vehicle's weight. For the covering abstract see ITRD E116488.
Abstract