Effects of stationary work zones on motorway crashes

The empirical Bayes (EB) before-and-after method was applied in this study in order to evaluate the change in the expected crash frequencies associated with the installation of work zones on motorways. A dataset of 15,570 stationary work zones including crash data, road segment data and traffic data was analyzed in order to estimate crash modification factors (CMFs) associated with different layout configurations of work zones. The findings of this research indicated a general increase in crash frequencies due to the installation of work zones. Furthermore the analysis has shown that all layout configurations that involve a crossover are very critical for safety: the highest CMFs are observed for the configurations with only a partial diversion of traffic to the opposite carriageway (with part of the traffic remaining in the ordinary flow direction and part travelling in the opposite carriageway).