Capacity of shared left turn lanes-comparative analysis

This paper compares the capacities of shared left turn lanes obtained by four methods for varying lane configurations, through and left turn volumes, and traffic signal timings. The 1997 Highway Capacity Manual (HCM), Canadian, SIDRA, and Levinson methods were analyzed for left turn volumes ranging up to 250 vehicles per lane per hour (vph) for both 60- and 90-s cycles, assuming 50% effective green time per cycle. Scenarios were tested assuming both equal and unequal volume. More than 700 individual computations were performed. The methods provide generally consistent patterns for each scenario tested. The 1997 HCM method (unlike the 1994 HCM method) consistently provided higher capacities than the other three methods for single-lane approaches. The SIDRA method consistently provided the lowest capacities on multilane approaches where left turn volumes exceed 100-150 vph. The research suggests further field tests and analyses to see if modifications in left turn equivalency factors for single-approach lanes associated with the 1997 HCM method are desirable to bring the results more in line with those of the other models. The general consistency of the shared lane capacities obtained by the Canadian and Levinson methods, along with their simplicity, clarity, and ease of use make them well suited, especially for quick response applications.