Estimating the Effect of Freeway Ramp Metering on Safety

This paper presents a method evaluating benefits of ramp metering strategies on freeway safety. Based on the traffic and the accident data collected on a 4.2 km (2.6 mile) section of Interstate highway 35-West in Minneapolis, U.S.A., the relationship between traffic variables and safety measures is investigated. An aggregate specification that could be used to predict accident frequencies on freeways is proposed as a multiple regression form. The specification includes 15 minutes volume and occupancy data, which are commonly available from surveillance and control systems. The primary variables that appear to affect the frequencies of freeway accident are: vehicle-miles of travel, entrance ramp volumes and the dynamic effect of queue building. A simulation method evaluating the dynamic effect of control strategies on safety is proposed next. The potential benefits of freeway ramp metering on freeway safety are finally investigated via a proposed method.