M. Jalayer, F. Baratian-Ghorghi, H. Zhou

Pages: 129-140

Abstract
Secondary crashes (SCs) are crashes resulting from non-recurring congestion and other unexpected surrounding conditions affected by a primary crash. This type of crash results in further disruptions and congestions. Previous studies showed that more than 20 percent of the primary crashes lead to SCs and the average duration of SCs is longer than the average duration of any other crashes, representing the necessity of further attempts. In order to identify the SCs, it is required to advance investigate their characteristics. However, there is a lack of available information to link a primary crash and a secondary crash, especially from the standpoint of time and location. This paper aims to develop a method to identify secondary crashes and then to study the statistical characteristics of this type of crash based on different criteria including, collision type, severity level, time of day, area type, average emergency response duration, and roadway classification from available historical crash databases and without the necessity of further data collection efforts. A proportional test was conducted to examine the statistically significant difference between the characteristics of the primary and secondary crashes. The study uses the latest available dataset (2010 to 2013) from the critical analysis reporting environment (CARE) to point out the secondary crash locations in the state of Alabama. The outcome of this study can help state and local agencies to identify SCs for performance measure of their incident management program and to develop countermeasures to prevent SCs.

Keywords: secondary crash; primary crash; highway incident management; hypothesis test; CARE; proportional test