In order to improve the public transport of Paris, the French capital authorities have decided the implementation of a new tramway line which itinerary will follow the so called Boulevards des Marchaux urban ring road. The present study is focused mainly the eastern area of the boulevards and particularly on the portion between Porte de Vincennes and Porte de Bagnolet. This tramway line will be physically separated from the boulevards road with exception of intersections. Urban planners considered 4 implementation design scenarios which are the axial, the bilateral and two variants of the lateral one. In order to get the best benefits from this new public facility, trams should not have to stop on signalized intersections. Consequently, tram pre-emption systems had to be designed, but respecting the fixed control plan concept which is used in the rest of the city intersections. This pre-emption system having a notable impact on the vicinity road network, the Paris city Council ordered a study aiming to compare the four implementation scenarios based on the following criteria, ranked in priority order: pedestrian safety, tramway commercial speed, capacity of the crossing streets and capacity of the boulevards. Analytical calculation based solutions are clearly not able to show the impacts of the pre-emption system applied on the 15 intersections considered on the study area. Therefore, the Paris City Council decided to undertake a traffic simulation study using the transport simulation package AIMSUN, and particularly its microsimulation tool. As mentioned before, the control plan has to be fixed when no tram is approaching the intersection. A fully adaptive control plan could not therefore be considered. Thus, when a tram call is received, the current phase has to be ended (respecting the minimum green time), an inter-phase to be then activated and finally the special tram phase to be set off. Note that this tram phase has to be activated before the tram reaches the braking area before the intersection. Once the tram exits the intersection (exit call) and when no other tram is on it, the control plan activate an inter-phase allowing to finally set off the next phase of the fixed plan. In a first step simulation based studies have been made relying only on one intersection of the network which has been considered as representative of the rest of intersections. For each scenario, there were different control plan options to be tested. Following the general objectives edited by the Paris City Council, their timing has been tuned to get the better from them. A final assessment aimed to select, for each scenario, the best control plan that will be applied for the whole network. In addition, a particular attention has been paid for the two signalised roundabouts included in the study area as the determination of the correct phases and timings has not been trivial. Indeed, problems of grid locks and vehicles blocked on the tramway platform had to be carefully studied in order to get the best options. Microsimulation of the whole tram corridor could then be run, and various outputs were used to assess the global performance of each scenario. The first output used was tram speed profiles in order to test if there was any deceleration not implied by tram stops. This allowed detecting any problem of vehicles getting trapped on the tramway platform or any malfunction of the pre-emption system. Capacity, or better said, queue length increases have been then measured at each point of the network in order to know where bottlenecks were and to evaluate the risk of congestion propagation that could lead to an intersection blockage. For each intersection, the upstream demands have been compared to the downstream throughput to get the total queue increase (in number of vehicles per hour). Finally, safety aspects have been analysed in term of numbers of potential conflict between pedestrians, bicycles (both were concretely simulated) and cars. In addition to that, non quantitative aspects have been highlighted, like the probability of non respecting red lights, but only thanks to the experience of the local engineers that were able to predict this type of events. The global evaluation, based on a multi criteria approach, finally showed that the axial scenario offered the best performances. Therefore, this implementation design will be selected for the Tramway des Marchaux (East side) for its construction. For the covering abstract see ITRD E137145.
Samenvatting