Freeway Incidents in the United States, United Kingdom, and Attica Tollway, Greece

Abstract

A summary of real-world incident statistics is presented from various freeways, and statistics from the Attica Tollway, in Greece, are simultaneously compared with incident data from U.S. and U.K. freeways. Detailed statistics are available from the Attica Tollway; statistics for U.S. and U.K. freeways were compiled from the available literature. Attica Tollway capacity under incident conditions is compared with estimates from the Highway Capacity Manual 2000. Interestingly, the incident capacity estimates by freeway number of lanes available and number of lanes closed because of an incident are practically identical for the Attica Tollway and U.S. freeways. This finding agrees with conclusions by others that similarly configured (standard) freeways produce similar incident profiles. The implication is that incident behavior could be predictable, but detailed analytical models are lacking. To this end, models are developed to explain the correlations between incident duration or queue length and a variety of independent variables such as number of crashed vehicles, number of responding vehicles, response times, number of lanes at the incident location, and various other temporal factors.