Performance Evaluation and Optimization of Two Closely Spaced Intersections: A Case Study of Ekantakuna and Nakkhu Dobato Intersections in Lalitpur

Abstract

Most of the intersections in Kathmandu Valley suffer from long queues and delays, especially during peak hour times of the day. Such long queues and delays are more pronounced in those intersections which are situated near to one another because of lack of coordination among the flow between the two intersections and also due to queue spillback from one intersection to the another. One such place is situated at Ekantakuna Lalitpur where two intersections viz. Ekantakuna Intersection and Nakkhu Dobato intersections are situated close to one another (~300m). Currently operating under manual traffic police control, these intersections face significant congestion due to high traffic volumes and terminal queuing interference. Using SIDRA INTERSECTION software, the existing traffic condition of the intersections were modelled to determine the existing Level of Service (LOS), Degree of Saturation (DOS) and Average delay at each of the intersections. It was found that Nakkhu Dobato intersection exhibited a global DOS of 2.33 (LOS F) while Ekantakuna reached a DOS of 1.63 (LOS E) during the peak hours. Then the software was used to optimize the intersections, individually with traffic signals and with coordinated traffic signal system to determine the improvement of the traffic flow through the intersections. The results showed that while isolated optimization provides marginal improvement, coordinated signal control yields the most significant improvements for the corridor. Under coordination, the global DOS at Ekantakuna improved to 0.98 (LOS D) and even though the Nakkhu Dobato intersection exhibited DOS of 1.7 (LOS D), it was far better than the original DOS at the intersection. These findings suggest that implementing coordinated signal systems is essential for managing the closely spaced intersections in oversaturated urban environments like the Kathmandu Valley.