AN INVESTIGATION OF THE IMPACTS OF RAMP METERING ON TRAFFIC FLOW WITH AND WITHOUT DIVERSION

The effect of various levels of ramp metering on traffic flow in a 7-mi-long urban corridor consisting of a freeway, two parallel surface arterials, and seven perpendicular connecting surface arterials was evaluated. The study was conducted both with and without traffic diversion from on-ramps to surface streets using the INTRAS freeway corridor simulation model. Three levels of ramp metering were analyzed to determine how much each would reduce the effects on traffic of an incident on the freeway. For each level of metering, several traffic diversion schemes were introduced in an incremental manner. The diversion of vehicles was implemented for each level of ramp metering until the remaining number of vehicles behind the meters was the same for each level and was less than the storage capacity of the ramp behind the meter. The main conclusion is that, whereas ramp metering improves the traffic flow on the freeway, it adversely affects the total system because of the overflow queues behind the meters, which spill back onto the surface streets. The only metering level that does not do this (in the absence of diversion) is one in which the metering rates are adjusted so that the overflow queues do not occur (equivalent to a queue detector at the upstream end of a ramp overriding the meter when a queue is detected). This level of metering, however, is rarely sufficient to overcome the capacity reduction resulting from an incident. To minimize the adverse effects of ramp metering, an appropriate traffic diversion plan for the implemented ramp metering strategy is required. The impacts on the total system under the optimum ramp metering and the best diversion plan consist of only a 4.1% increase in speed and a 10.5% decrease in delay.