Performance Evaluation Procedure for Bus Interchange Areas

Performance Evaluation Procedure for Bus Interchange Areas


Josep Mension, TMB, Miquel Salicrú, University of Barcelona, Miquel, Estrada


This paper aims at defining the methodological framework for analyzing the bus service performance at interchange areas.


Several cities show simple and readable bus network configurations (New York, Barcelona) rather than circuitous, door-to-door services. The former concept is usually based on a quasi-perfect grid of bus corridors or a radial scheme. It allows a significant reduction of resources as well as a higher demand coming from non-recurrent passengers due to the ease of network interpretation. Nevertheless, these simple bus configurations are operated by transfer-based services. It means that there is a small part of direct trips in the system and the vast majority of passengers must do a transferring operation between different routes. Several scientific contributions demonstrated that transfer-based networks provide user travel time savings compared to door-to-door services if bus cruising speed is improved and the transfer operation is made at a reasonable time and cost.
In fact, transfer stations or interchanges areas have become the cornerstone of these systems. However, little attention has been devoted to the quality and performance of bus transfer operations. Traditionally, the synchronization of outbound and inbound services has been proposed for metropolitan routes with medium-high headways. Unfortunately, for high performance bus systems in urban areas with short headways, the proposed metrics to control the performance operation are provided at a route or network scale. There is a lack of contributions analyzing the proper balance between the spatial configuration of all stops in the interchange area and the temporal coordination of arrival-departures.
This paper aims at defining the methodological framework for analyzing the bus service performance at interchange areas (IA). A transfer quality ratio (TQR) is proposed as an evaluating metric to characterize the bus service quality in interchange areas. Barabino et al (2012) stated that the user satisfaction should consider the overall quality service dimensions (planning, target service adherence, reliability, information, safety, security, cleanliness). In this way, the operating-side attributes of interchange areas are related to the time needed to complete a trip between two routes in the IA. This time depends on i) the spatial distribution of stops over the IA (it defines the walking time component) and ii) the temporal coordination between arrivals and departures (it defines the waiting time at the outbound route).
If we calculate the average time needed to complete one trip in the IA (mu) and its standard deviation (sigma), we may present the user satisfaction function with regard to operating issues by the form S(muT, sigmaT):R^2 → R. Although we do not know in advance the analytical formulation of S(muT,sigmaT), we can state that S must be a regular function (second derivative is a continuous function) and must present a maximal value at (mu0, sigma0). This paper statistically demonstrates that the assessment surfaces (TQRo,muTsigmaT) are defined by ellipsoids and the locus of points (muT, sigmaT) that presents the same user satisfaction is defined by ellipses. These ellipses should be rotated a given angle to control the available TQR domain. The contour conditions and other required constraints for generating these assessment ellipses are presented.
Therefore, a consistent evaluation methodology based on these statistically-developed ellipses is presented to rank the provision of interchange areas. This methodology is suitable for any general surface Interchange areas operated by transit modes at short-medium headways. This methodology has been used to evaluate four interchange areas of the new bus network of Barcelona. This transfer-based network has been implemented in phases from 2013. The master plan will consist of 25 grid corridors+3 diagonal-shaped bus routes. At this moment, there are only 16 bus routes in service, with 51 interchange areas. The spatial distribution of the four selected IAs is considered to be constant, and therefore only causes a translation of the ellipse in the axis muT Results show that the regularity of outbound routes strongly affects the performance of these transfer stations. The transit states (muT,sigmaT) usually fall in an acceptable performance domain, but there are some (muT,sigmaT) pairs associated to a low level of service. These pairs correspond to time intervals of irregular bus arrival at stops. Finally, we identify the existence of a statistic correlation between muT and sigmaT variables


Association for European Transport