Introduction

Changing trains is a crucial part of many rail journeys. It is essential for reaching all destinations, but it takes time and is often stressful, especially when the arriving train is delayed. In models for cost-benefit analyses, interchange time for the railway is typically considered 2-3 times more expensive than in-vehicle time, see, e.g., ASEK (Trafikverket, 2020). Designing and operating railway stations in a way that could minimise the required interchange time is therefore vital for making rail travel attractive.

In this paper we address the problem of using the platform tracks in the best way, balancing crossing train paths with easy interchanges across the same platform. Primarily, we look at the problem on a tactical level, i.e., timetable planning, but the insights are also useful on a strategic level for designing and re-designing railway stations, and on an operational level, doing the dispatching. This work is inspired by Johansson and Nilsson (2021), who have made an initial simulation study of various platform allocation strategies using the Arena Simulation Software for general event-based simulation.

Most of the models proposed for the train platforming problem focus on feasibility, i.e., to find a conflict-free solution, where no two trains are assigned to the same resource at the same time. The objective is less important, and typically just expresses some cost or weight for each train and/or platform. Carey and Carville (2003) for example, define for each pair of train and platform a “platform obstruction cost”. Lusby et al. (2011) have made an overview of models and methods for railway track allocation, which also includes platforming and discusses the aspect on a strategic, tactical and operational level. While the previously proposed models typically focus on feasibility, this study aims for a platform allocation that balances the need for short interchange time for the passengers across the same platform towards the risk of knock-on delays when two trains are forced to use the same infrastructural resources.

Method

Train platforming is a well-known problem and optimising the platform assignment while accounting for train changes can be approached by optimisation models, e.g. as done by Calderon (2022). In this paper, however, we aim at exploring principal differences when prioritising a few crossing train paths versus fast interchanges. To highlight the difference, we will use some very simple, ideal principles for allocating trains to platform tracks. These principles are using the track allocation from a given timetable, using the lowest numbered track that is free, and maximising train changes at the same platform.

The analysis is performed for the medium-sized station Norrköping, Sweden, using the timetable of a normal Thursday in autumn 2016. In the microscopic railway simulation tool RailSys, the infrastructure as well as the timetable is modelled in detail (Bendfeldt et al., 2000). At stations, each train has a specific platform track allocated, including the detailed train path through the station. To investigate track allocation changes and their effects, two platform reschedulings have been performed in RailSys and studied by counting the resulting number of crossing train paths and changes at the same platform. For this study, possible connections between trains are defined as when there is a minimum of 5 minutes and a maximum of 30 minutes between the arrival time of a train and the departure time of another train. If the time is less than 5 minutes, the time is too short for changing trains, and if more than 30 minutes, we assume that the track allocation is of less importance to the passengers, who might use the time to leave the platform for station building shops and facilities.

When counting the number of possible connections as described, it is also observed how many of these train pairs have crossing train paths, i.e. to some extent use the same infrastructure at the station, and how many of the possible train changes that would take place at the same platform, either across the same platform or to a later train arriving at the same track. Furthermore, the resulting capacity utilisation in per cent for the considered time periods of the different alternatives has been calculated with the method introduced by Weik et al. (2020). This method looks at the whole station, from the entry signal to the exit signal, and performs a timetable compression based on the standard by UIC (2013) for a given time period. The exact block occupation times and conflicts between trains regarding infrastructure usage are accounted for in the compression.

Analysis and results

During the full day, 205 passenger trains from both directions were routed through Norrköping C. The total number of possible connections between these passenger trains was 1,033. The number of crossing train paths as well as the capacity utilisation increases when the number of connections from the same platform increases. Future work should, e.g., model connections more accurately, perform RailSys simulation of each track allocation to assess the spread of delays, and apply cost-benefit analysis to find the best balance between the conflicting goals of having few crossing train paths and many train changes at the same platform.

References

Bendfeldt, J.P., Mohr, U., Müller, L., 2000. RailSys, a system to plan future railway needs. Comput. Railw. VII 50, 249–255.

Calderon, S., 2022. An ILP-model for the Train platforming problem (Master’s thesis). Linköping University.

Carey, M., Carville, S., 2003. Scheduling and platforming trains at busy complex stations. Transp. Res. Part Policy Pract. 37, 195–224. https://doi.org/10.1016/S0965-8564(02)00012-5

Johansson, E., Nilsson, H., 2021. Station capacity and platform allocation – a test case at Linköping central station (Master’s thesis). Linköping University.

Lusby, R.M., Larsen, J., Ehrgott, M., Ryan, D., 2011. Railway track allocation: models and methods. Spectr. 33, 843–883. https://doi.org/10.1007/s00291-009-0189-0

Trafikverket, 2020. Analysmetod och samhällsekonomiska kalkylvärden för transportsektorn: ASEK 7.0 (Report).

UIC, 2013. Code 406 - capacity. International Union of Railways, 2nd edition, Paris, France.

Weik, N., Warg, J., Johansson, I., Bohlin, M., Nießen, N., 2020. Extending UIC 406-based capacity analysis – New approaches for railway nodes and network effects. J. Rail Transp. Plan. Manag. 15. https://doi.org/10.1016/j.jrtpm.2020.100199