How and when cordon metering can reduce travel times

Jean Doig,Carlos F. Daganzo,Michael J. Cassidy
DOI: https://doi.org/10.1016/j.trc.2024.104581
IF: 9.022
2024-04-12
Transportation Research Part C Emerging Technologies
Abstract:The paper addresses two questions regarding cordon metering that have until now gone unanswered. The first of these pertains to how and where a metered cordon ought to be placed in a city to be of greatest benefit. A simple 3-step rule is proposed that can be readily applied in real settings, and that we call the cordon layout conjecture, or CLC. Its use is shown to minimize the overall travel time inside and outside the cordon combined. The second question pertains to the conditions for which an optimally placed metered cordon can reduce said travel time relative to doing nothing. The answer to this question may be disappointing to some readers. We find that metering can reduce this travel time only for congestion levels that well exceed what we observe in real cities of the world. Of course, cordon metering might still have its place, e.g., to transfer congestion from downtown areas with extensive bus operations to outlying areas without buses; or perhaps to curb demand for car travel. The CLC could help in these endeavors. Findings came by simulating traffic in regions with three geographic configurations. They are: an isolated city center within which all trips begin and end; the same city center with suburbs that bring traffic into the city via arterials; and the city center with suburbs and exurbs that dump traffic into the city via freeway off-ramps. These configurations were studied under multiple demand scenarios, including those that generated severe congestion. Parametric tests for a wide range of cordon placements support the CLC. Optimal placements often turned out to reside well inside the city center's periphery, which is distinct from what has been tried in the literature. Findings also show how suboptimal cordon placements can do damage. And they show how CLC-placed cordons can reduce overall travel time (by close to 15% in some cases), though again, only when congestion levels far exceed anything that we could measure in real cities.
transportation science & technology
What problem does this paper attempt to address?