Saturday , 21 September 2019

Testing the performance of Displaced Right-Turn Intersection in mixed traffic conditions

Studies on Displaced Right-Turn Intersection design show that it is a more effective design for managing traffic at intersections instead of the conventional 4-arm design

The rapid growth of road traffic in Indian cities is making the available transport infrastructure inadequate. There is a need to find alternative solutions to the problem. One such method is the Displaced Right-Turn Intersection design. The concept of the displaced right turn has been considered since the 1950s as an alternative intersection design to traditional at-grade and grade-separated intersections. It enables one or more conflicting movements to take place away from the main intersection at a new crossover intersection which reduces the number of conflicts at the central node. This concept has proved effective in the developed countries (UK, USA). The need for the present study is to determine whether the Displaced Right Turn (DRT) design is effective for highly heterogeneous traffic flow conditions prevailing in India.

Displaced Right-Turn Intersection

Displaced Right-Turn intersection is an at-grade signalised intersection which can be considered as a sub level between grade separated and at-grade intersections. The principle behind it is that the main conflicting traffic, i.e., the right-turning traffic gets excluded from the main intersection and is carried to a bay which runs parallel to the opposite through-traffic. This right-turning traffic is carried through a crossover intersection approximately 300-400 feet upstream of the main intersection. The main advantage of this intersection is that for a partial DRT (DRT on two major road approaches), there is a single phase reduction resulting in only three phases for a 4-arm intersection. A double phase reduction can be achieved using a complete DRT (DRT on all four approaches).
The major advantages of DRT are listed below:

  • As said earlier, there is a one phase reduction for partial DRT and a two phase reduction for a complete DRT. This traffic signal phase reduction in DRT results in less delay, fewer pollutants and lower fuel consumption than the conventional 4-Arm intersection.
  • The intersection can easily handle more volume and is efficient in terms of land use. Thus, it avoids the construction of grade-separated intersections at a later stage.
  • The number of conflicting points in DRT is 28 while in the case of 4-Arm intersection, it is 32. When the number of crossing conflict points (which accounts for higher risk) is just 12 in the case of DRT, it is 16 in the case of ordinary 4-Arm. It shows that DRT may also provide a certain level of safety benefit compared to the ordinary 4-Arm intersection.

Field Requirements

Complete DRT/PDRT

  • Extra right-of-way in all approaches
  • Two traffic signals at each crossover junction
  • Traffic signal at the main intersection where the right-turn parallel bay meets the main intersection

Phasing Diagram for DRT

In an ordinary 4-Arm intersection, a minimum of four phases are required to allow the flow of traffic while in a DRT we require only two, and in the case of partial DRT, we require only three. Figures 1(a), (b) show the phases for a complete DRT and figures 2(a), (b), (c), show the phases for a partial DRT. (See next page)

  1. Complete DRT


figure 3 : Study Intersection - Bhagwan Mahavir Chowk

In phase 1 of DRT, the N-S through moving traffic is allowed. At the same time, the right turning traffic from the parallel bay is also allowed. This right turning traffic then faces a signal at the crossover junction on E-W road. The right turning traffic of the E-W road is now taken to the parallel bay and it stops at the main intersection. The entire left turning traffic is free to move and is unsignalised in both the phases


2.   Partial DRT

In phase 2 of DRT, the E-W through moving traffic is allowed. This phase is identical to phase 1 except that it is with respect to the E-W direction. This completes the cycle, thus giving rise to just two phases to discharge all the traffic flows.
Phase 1 of a Partial DRT (PDRT) is identical to phase 1 of DRT. During phase 2 and phase 3, the East-bound and West-bound traffic is cleared along with its right turn traffic. When this happens, the right-turning traffic of the North-bound or South-bound traffic (whichever is free to cross) is allowed to cross over to the parallel bay, and is queued at the main intersection.

Need for study

The conventional intersection designs fail to control the ever-increasing traffic at saturated traffic flow conditions. Grade-separated intersections are very costly and are also space-demanding. Therefore, such measures often face obstacles. This brings us to the Displaced Right-Turn intersection. The DRT/PDRT has been proved to be effective in the developed countries (UK, USA). The need for the present study is to determine whether the PDRT design is effective for highly heterogeneous traffic flow conditions prevailing in India. The reason for selecting PDRT instead of DRT is that DRT demands more land space which is usually unavailable in India. Also since the concept is new to a highly heterogeneous traffic, it is feasible to begin the research with PDRT. The idea is to determine whether the design itself is more efficient at the intersection level. The main objective of this study is to evaluate the impact of Partial Displaced Right-Turn Intersection under heterogeneous traffic flow conditions and erratic driving behaviour for various traffic volume and turning proportions.

Data collection and extraction

The Bhagwan Mahavir Chowk in Vashi, Navi Mumbai (Figure 3) is a 4-Arm junction and carries medium-heavy traffic flow. As shown in Figure 3, areas 1 and 2 allow us to design a partial DRT in this 4-Arm intersection and thereby, improve the network system. The approaches A and C are major 2-Lane approaches, while B is a semi major 2-Lane road. Traffic D is a minor 2-Lane approach. A video-graphic survey of the intersection was performed for about one hour and data extraction was done in the lab. Field observed traffic flow details and traffic composition are given below in Table 1.

Model development and validation

Model Validation is considered to be the process of determining to what extent the model’s underlying fundamental rules and relationships are able to adequately capture the targeted emergent behaviour, as specified within the relevant theory and as demonstrated by field data. In other words, validation is the overall process of comparing the model and its behaviour with the real system and its behaviour. A VISSIM model was made to simulate these conditions and validate the model for further analysis. Simulation run time was set as 3600 seconds to replicate the real conditions. The validation criteria, as obtained from the field, were as follows: Total number of vehicles in the network during the period of survey: 5017 and average queue length in approach A: 56m.


Figure 4: VISSIM model snapshot of 4-Arm conventional Intersection - Bhagwan Mahavir Chowk

The total number of vehicles in the network as obtained for the data collection points provided in the 4-Arms was found to be 4775. We obtain an error of just 4.82% in this case. The simulated result for the queue in approach arm A was 60m while the observed queue length was 56m. Therefore, the error is 7.14%. Since both the validation criteria are met with relatively high accuracy, the VISSIM model of Bhagwan Mahavir Chowk can be regarded as a true representation of the real system. Therefore, the model can be used for further analysis and experimentation.

Effect of Partial DRT intersection

For the application and analysis of PDRT, first of all a comparison has to be made with the conventional system. Therefore, the 4-Arm VISSIM model is simulated for the various conditions and the same conditions are applied for the 4-Arm PDRT model. The snapshots of developed models of the study intersection are shown in figures 4 and 5.

Performance of conventional 4-arm intersection & PDRT intersection


Figure 4: VISSIM model snapshot of 4-Arm PDRT Intersection – Bhagwan Mahavir Chowk

The VISSIM model of the Bhagwan Mahavir Chowk was considered to be the base model representing the existing condition. It was then tested for the traffic volumes 1500, 2000, 2500 and 3000 vehicles per hour for the right-turning proportions of 10%, 20%, 30% and 40%. These variations were made only in the major arms vis-a-vis arms ‘A’ and ‘C’. The model results obtained were from the simulation.
The VISSIM model of Bhagwan Mahavir Chowk was redesigned into a 4-arm PDRT intersection – 2 lanes per approach with the design implemented on the arms A and C. The model was then tested for the same traffic conditions. Another VISSIM model was designed into a PDRT intersection with the design implemented on the arms A and C. But in this one, instead of 2-Lanes per approach in the major arms, it was given 3-Lanes per approach. The model was then tested for the same traffic conditions.


Analysis of Simulation Result

The simulation results of PDRT 2-Lane and PDRT 3-Lane were compared with that of the existing condition. The following parametres were considered for the analysis: (i) Percentage reduction in Average Delay Time per vehicle (Table 2) (ii) Percentage reduction in the Average Queue Length in arms A and C (Table 3)

Major findings

  • Overall, the PDRT intersection is an effective way of improving traffic flow at an intersection. But for 2-Lane PDRT with more than 20% right turning traffic, the system isn’t efficient enough.
  • There is a considerable reduction in average delay time and average queue length, which are two major parametres that help us judge a system’s efficiency.
  • In the case of 2-Lane PDRT, the queues of right turning vehicles block the through movement traffic. Therefore, it is necessary to provide adequate signal time and phases to reduce the queues at the crossover intersection. In the case of 3-Lane PDRT, this conflict can be avoided by providing the 3rd lane as a right-turning bay and by giving pavement markers and other road signs to indicate the right-turn intersection about 100m upstream.
  • In practice it is found that for most of the major approaches, the through movement traffic will be relatively higher in number compared to the right or left turning traffic. The PDRT intersection particularly enhances this movement by allowing the through movements of both the arms of the major road simultaneously. This implies more green time and lesser conflict for this volume, thus reducing the delay times and queue lengths.

PDRT is more effective

The analysis shows that the PDRT design is certainly more effective when compared with the conventional 4-arm design. The new design offers lesser delay time and queue length, supplementing it with a higher handling capacity and thereby, almost removing the necessity for a grade separated intersection. Thus, the intersection falls into a sub-level between the grade separated and at-grade intersections, and can be regarded as a measure of sustainable development in transportation engineering. The study has proved that the design is certainly effective at the intersection design level. Since the analysis has shown positive and promising results, further detailed analysis can be extended to a corridor level and research may also be done for complete DRT designs.




Dr P Vedagiri
Assistant Professor
Transportation Systems Engineering
Department of Civil Engineering,
Indian Institute of Technology Bombay, Mumbai, India

Mr Krishnakumar Subramanian
3rd year, B.Tech Civil Engineering,
National Institute of Technology, Calicut

Sign up to see more

By continuing, you agree to privacy policy