Civil Engineering May 2022 | Vol 30 No 4

40 May 2022 Civil Engineering We used drones to observe driver behaviour during peak traffic times and found instances of drivers using turning lanes and shoulders to bypass long queues at intersections. For instance, Figure 1 illustrates a MBT on a through-movement driving in the right-turn lane, which has a shorter queue than the through lanes. After the traffic signal turns green, the taxi is seen cutting back into the adjacent lane, thereby effectively skipping eight vehicles in the queue and saving about 24 seconds of delay. Such driving is of course problematic from the perspectives of safety and orderly road use, but it illustrates the pressure on drivers to increase operating speeds in order to reduce route cycle times and attract more passengers. Can we use dedicated infrastructure to formalise this behaviour and help achieve the same ben- efits while reducing the downsides? We conducted three simulation experiments to start exploring this question. SIMULATIONS 1. Minibus taxi priority: Is there a case for priority infrastructure for informal vehicles only? In this study, a macroscopic traffic model was developed of a hypothetical intersec- tion with four potential priority treat- ments: a regular curbside stop (as the base case), a queue-jumping lane for MBTs only, a single lane pre-signal strategy, and a continuous taxi lane (Figure 2). Queue-jumping lanes are short lanes that allow MBTs to bypass queues at congested intersections; single-lane pre-signal strategies do the same but use the oncoming traffic lane temporarily in the reverse direction for this purpose. We estimated the costs and benefits of each treatment to taxis, other vehicles, and the road agency under a range of conditions including various flow rates under typical urban conditions. The findings showed that: Q Q While costs and benefits vary substan- tially across the different treatments, all of these priority measures offer the possibility of substantial reductions in social (total) costs. Q Q Both taxi operators and passengers benefit substantially: estimated savings for operators (in fuel reductions and efficiency gains) are between R1 100 and R9 000 per month (Table 1), which translates into an implicit subsidy to the MBT sector that can promote sustain- ability and affordability in the industry. Q Q At low to medium volumes of general traffic, there is little impact on car users. Q Q Queue-jumping lanes are the most preferred solution due to their rela- tively low implementation cost. 2. Hybrid lanes: Can we use BRT lanes better as multi-class public transport lanes? This study explored the question of whether we should rethink the exclusive use of BRT lanes by BRT buses. The logic is that some BRT trunk routes in South African cities have such low bus volumes that they are completely underutilised. Could excess capacity in bus lanes be used to accommodate informal vehicles in a hybrid system, without substantially degrading the service offered to either bus or minibus passengers? We developed a detailed micro-simu- lation model of a 1 200 m-long section of the A Re Yeng BRT corridor along Nana Sita Street in Tshwane, asking what would happen if MBTs were allowed to use the BRT lane (Figure 3). The results show that, under uncon- gested conditions, there is little benefit gained from such hybrid operations. However, as congestion sets in during peak hours, a clear case can be made for allowing MBTs to share bus lanes under specific conditions. Even at the modest MBT volumes currently seen in this cor- ridor (up to 94 vehicles per hour per di- rection), MBTs contribute significantly to congestion in general lanes, and shifting them to bus lanes reduces demand to below capacity, leading to significant sav- ings for private vehicle users. Taxi passengers benefit from an up to 50% reduction in travel time, with minimal impact on buses. The bus lane, which currently carries between two and nine buses per hour per direction, can accom- modate an additional 87 buses per hour or 334 MBTs per hour before service deterio- rates to unacceptable levels at stations or intersections. These results persist even when taxi and bus volumes are increased to take account of modest demand growth and latent demand, although the rules of sharing infrastructure become critical. These results are not necessarily gen- eralisable as they depend on the specific taxi stopping patterns and the geometric conditions (specifically the presence of overtaking lanes) in this corridor. We do not advocate turning existing BRT lanes into hybrid lanes without carefully considering issues of station design, en- forcement, and passenger safety. But, the results do illustrate that cases exist where dedicated infrastructure may work better as multi-class public transport lanes, rather than exclusive bus lanes. We might have to think of a step- wise approach where, for a period of Curb-side stop Single lane pre- signal strategy Queue- jumping lane Continuous taxi lane Figure 2 Four alternatives for MBT priority at intersections Table 1 Example of financial impacts of priority infrastructure on MBT operators Infrastructure Hourly taxi operating cost Operating cost savings/ taxi Minimum monthly savings/taxi Maximum monthly savings/taxi Curb-side taxi stop R133 – – – Queue-jumping lane R105 R28 R1 232 R4 928 Single lane pre-signal strategy R108 R25 R1 100 R4 400 Dedicated taxi lane R82 R51 R2 244 R8 976