Assessment of unprotected left-turn capacity models using arena simulation

Gap acceptance theory based on the stochastic process is commonly used to analyze unprotected left-turn capacity at two-way stop-controlled intersections or signalized intersections with a permitted left-turn phase. However, there are some factors in the current HCM models that are found to confound the stochastic process. This paper presents new models to improve capacity estimation for the unprotected left turn. According to the simulation results using the Arena models, it was found that for TWSC intersections, if there are a relatively higher number of heavy vehicles in the left-turn traffic, the HCM model may underestimate the left-turn capacity. For signalized intersections, the HCM model tends to under-estimate the unprotected left-turn capacity for the lower opposing-through traffic volume and over-estimate it for the higher volume. The heavy vehicles in the left-turn traffic can increase the under-estimation trend. Furthermore, the simulation results verified that the proposed new models are reliable.