Modelling fire occurrences in heavy goods vehicles in road tunnels

Abstract

The project reported in this paper has been organized to scrutinize current incident data on near fires and fully developed fires in Norwegian road tunnels longer than 500 m. This length is chosen because it is assumed that shorter tunnels are less critical in case of fires. The project included collecting data and transferring it into formats enabling mathematical modelling. The major issue of this work has been to resolve: What are the major contributing tunnel infrastructure factors leading to heavy goods vehicle (HGV) fires in Norwegian tunnels? By using Poisson regression modelling, several models are developed showing good fit with the observations. All models reveal that slope, length, annual average daily traffic of heavy goods vehicles, and whether a tunnel is subsea are significant factors. The most important is the subsea factor, and the effect of other risk factors is also more severe for subsea tunnels. The work also discusses weaknesses in the data material and the fact that there are several other interesting factors, for example related to the state of HGVs and driver behavior that are currently missing. The research potential for better modelling and understanding of HGV fires in tunnels is huge.