A Model and Case Study of the Impacts of Stochastic Capacity on Freeway Traffic Flow Benefits and Costs

When freeway traffic flow approaches capacity, minor disturbances or perturbations can cause unstable traffic streams to break down into queued or bottleneck conditions with the accompanying heavy congestion costs. As the traffic volume at which flow breaks down is uncertain, this research utilizes a stochastic capacity model to estimate congestion costs in terms of delays, fuel, and emissions. We apply this stochastic model to a congested freeway corridor in Portland, Oregon in order to demonstrate the impact of various traffic parameters on the net social benefits of traffic flow. Travel time is the dominant cost, followed by fuel costs. For a given value per trip (in $/mile), the traffic flow volume that maximizes social benefits decreases as travel time reliability decreases. Traffic flows near capacity levels are justified by trip values that are 50% higher if the impacts of stochastic freeway capacity are considered. Comparing macroscopic peak-period traffic characteristics among urban areas of varying size and density, we see that for a given peak-period trip value denser urban areas will have higher optimal flow rates. This comparison demonstrates a fundamental cost trade-off for traffic networks between shorter trip lengths and higher traffic intensity from increased urban density.