Development of proposed priority infrastructure for minibus-taxis and peak hour traffic count data

The paratransit industry in South Africa which mainly includes the minibus-taxis is growing at a fast pace. Thus, it has become the largest mobility supplier to the urban public. In Gauteng province, the economic hub of South Africa that includes Johannesburg, Tshwane and Ekurhuleni, minibus-taxis account for 46% of all peak-period passenger trips followed by private cars accounting for 44%, while buses and trains account for a combined total of 10% of peak-period. Unlike buses which have seen the provision of priority infrastructure at intersections in the form of bus rapid transit (BRT) with priority transit signals (PTS) to improve their efficiency, minibus-taxis currently do not enjoy the same benefits. However, any efforts of road authorities in South Africa to consider incorporating priority infrastructure for minibus-taxis would be constrained by the absence of literature suggesting the ideal choices and the design analytical procedures.

This research study aims to develop and evaluate design strategies for priority infrastructure for minibus-taxis at signalised intersections. Priority infrastructure at intersections can be in form of roadway facility infrastructure such as queue-jumping lanes, shared traffic lane, exclusive lanes or can be implemented via signal control. These infrastructure types are designed to provide efficiency benefits to road users mainly public transport such as buses. The first objective of this study is to develop an approach for identifying the design strategies for priority infrastructure for minibus-taxis at signalised intersections. A qualitative data method utilising document analysis technique is used to develop a framework matrix table to show the relationship between the geometric elements and the design treatments of priority infrastructure. Two categories of minibus-taxis (MBT) design strategies are then formed: 1) design strategies that only require repurposing of the existing intersection, 2) the design strategies that require major geometric improvements.

Secondly, an analytical approach is developed to evaluate the performance of two proposed design strategies using real world traffic data. To begin with, four isolated intersections in the city of Tshwane are evaluated for feasibility of the MBT design strategies. The framework matrix analysis developed earlier is utilised to select and evaluate the design strategies associated with the four intersections. In addition, the intersections are further assessed for safety, traffic operations and cost effectiveness. Eventually, the two most effective design strategies are selected for a detailed performance evaluation: 1) a shared MBT lane to be used by through movement minibus-taxis and left-turning vehicles (MBT+LT) and 2) a dedicated MBT lane for through minibus-taxis only. The approach uses modified analytical principles from the Highway Capacity Manual (HCM) to measure the performance of the selected design strategies using peak hour traffic data. The performance measures include volume to capacity ratio (v/c ratio), average vehicle delay, and adequacy of storage length of MBT priority lanes. The performances of existing intersections are compared with the performances of intersections after implementing the MBT design strategies. In general, the results show that the two proposed MBT design strategies significantly improved the performance of minibus-taxis at intersections while slightly reducing the performance of traffic in non-priority lanes.

Lastly, using the results from the two evaluated design strategies, a sensitivity analysis is performed on the modified HCM method to determine a range of traffic volumes for which the selected design strategies are feasible. Consequently, two models are set using a modified HCM method to evaluate two typical MBT design strategies involving a shared MBT lane and a dedicated MBT lane. The models are set to measure the v/c ratios of individual lanes on the approach as a measure of performances. The models are set to measure the highest v/c ratios while varying the traffic volumes at constant values of g/C ratios. The model outputs are in the form of graphs showing the relationship between left turning (LT) traffic, straight (MBT+T) traffic and v/c ratios at constant values of g/C ratios. These charts are developed as a planning and design guide when evaluating the feasibility of signalised intersections for the two evaluated MBT priority infrastructure types.

Overall, the study provides the first detailed results supporting the viability of priority infrastructure for minibus-taxis at signalised intersections. It also gives a detailed methodology and steps that could be used by traffic engineers and planners to design and evaluate the performance of priority infrastructure for minibus-taxis at signalised intersections. The matrix framework method and graphs for traffic volumes could provide planners with a structured way to identify feasible designs for the priority infrastructure for minibus-taxis at signalised intersections. The methodology used in this study can be adopted to evaluate other types of design strategies not evaluated in this study.

The study concludes that with well optimised design solutions, it is possible to use priority infrastructure to improve the performance of minibus-taxis at signalised intersections without adversely affecting the performance of traffic in the non-priority lanes.