Guidance for Simulation-Based Modeling of Auxiliary through Lanes

Abstract

Auxiliary through lanes (ATLs) are a cost-effective means for reducing congestion at signalized intersections, but little guidance is currently available on how to model ATL operations in a microsimulation environment. Without calibration, microsimulation tools assign some vehicles to an ATL as a function of basic lane-changing algorithms; this process results in much higher ATL utilization than observed in the field. This paper presents empirical models that explain ATL use as a function of through-movement congestion. This level of ATL utilization can be achieved in simulation tools by altering certain parameters, chiefly the upstream decision distance (UDD) of the ATL. Peak-hour counts and signal-timing data were collected at 22 ATL approaches from across the United States and coded into simulation to compare field and simulated ATL utilization. Modification of the UDD enabled close modeling of ATL utilization at 19 of 22 sites; the remaining three ATLs could not be modeled accurately in simulation because they experienced extremely low field utilization (less than 10% of total through-movement flow). After calibration, a model that approximated the correct UDD as a function of the total entering through-movement flow and ATL length was empirically developed. This model will allow practitioners to have a starting point when modeling a planned ATL in microsimulation packages; practitioners can then use the ATL operational models for further calibration of the UDD to simulate the correct level of ATL use.