Comparative Performance Evaluation of Roundabout Three-Legged and Four-Legged Intersection Insights from SIDRA and Statistical Modelling

Abstract

This study examined the operational performance of eleven urban intersections by combining field observations with SIDRA outputs and statistical analysis. The assessment focused on three common intersection configurations—roundabouts, four-leg, and three- leg layouts—using control delay, queue length, and Level of Service (LOS) as key performance indicators. Findings show that roundabouts consistently outperformed the other types, recording average delays of about 35–57 seconds and LOS grades ranging from C to E, which reflect generally stable traffic flow. Four-leg intersections demonstrated only moderate efficiency, with delays between 55–80 seconds and LOS levels of D to E. In contrast, three-leg intersections performed the poorest, exhibiting delays of 90–134 seconds and LOS F, signalling severe congestion and oversaturation. Analysis using ANOVA and Tukey’s post hoc tests demonstrated that performance differences among the three intersection types were statistically significant (p < 0.001), highlighting the critical influence of geometric design on traffic operations. Further regression analysis indicated that variations in delay and queue lengths were largely determined by intersection configuration rather than by traffic volumes or the number of lanes (R² > 0.88). These findings underscore the operational advantages of roundabouts under mixed-traffic conditions and emphasize the importance of implementing geometric and control improvements at both three- and four-legged intersections. The study advances understanding of intersection performance in rapidly growing urban settings and offers practical, evidence-based guidance for enhancing traffic management and network efficiency.