HIGH-FIDELITY MICROSCOPIC TRAFFIC SIMULATION FOR TOLL AND CONGESTION PRICING MODELING
Ramachandran Balakrishna, Caliper Corporation
We discuss a microscopic traffic simulation framework uniquely suited to the modeling and objective evaluation of tolling and congestion pricing scenarios at the network level. The high-fidelity approach facilitates the flexible assessment of any number of road user and vehicle classes, together with a highly detailed traffic simulation environment at the lane level. We illustrate the methodology through sensitivity tests on the Miami, Florida network. We also describe the real-world application of the same by a private toll-road operator tasked with guaranteeing a minimum rush-hour level of service on the toll facility.
The increasing difficulty of building new roadway capacity in dense urban areas highlights the need to better manage existing transportation infrastructure. While this problem typically benefits from a multi-modal treatment, tolls and pricing offer a mechanism to improve level of service on select highway facilities while simultaneously requiring drivers to pay for the externalities they impose on fellow drivers. Pricing is increasingly perceived as a valuable congestion mitigation tool in practice. Indeed, cordon pricing and high-occupancy toll (HOT) lanes are rapidly being deployed globally. London’s congestion charging scheme provides a high-profile example from Europe, while the US has seen a similar interest in such measures with operational or proposed systems in Washington, D.C., Miami (Florida), Seattle (Washington) and Southern California among others.
The impact of tolling/pricing strategies depends on drivers’ decision-making when faced with these options, and the transportation network’s aggregate response to those individual decisions. Level of service in turn advices drivers who must trade off the monetary cost against their willingness to pay. The economics of roadway pricing is backed by established research and empirical analyses focusing on user benefits. However, the evaluation of tolling strategies at the network level calls for the analysis of complex, dynamic demand-supply interactions. The demand side, for example, is driven by temporal demand profiles, value of time distributions, drivers' lane choice decisions, the mix of trip purposes, etc. The supply side is governed by vehicle mix, driving behavior, physical toll infrastructure configurations, and drivers' familiarity with the same. The implementation of pricing schemes must also consider network impacts on non-toll facilities and other transportation modes, from the perspectives of level of service and equity. Traffic simulation offers a flexible test-bed to estimate these impacts and benefits before deployment.
In this paper, we discuss a high-fidelity microscopic traffic simulation framework with unique capabilities in modeling and evaluating diverse tolling methods. We describe the real-world use of the approach by a private toll operator to compute the level of service and revenue impacts of various toll pricing algorithms and settings under different demand and supply conditions. These tests were performed for the deployment of pricing on the I-495 highway outside Washington, D. C. We also present a case study undertaken on the I-95 corridor in Miami, Florida as a practical demonstration of toll benefit and cost comparisons through traffic simulation. A variety of scenarios are analyzed, including origin-destination tolls, cordon tolls, HOT lanes and congestion pricing.
Association for European Transport