Network Effects in Railway Systems

Network Effects in Railway Systems


A Landex, O Anker Nielsen, CTT, Technical University of Denmark, DK


Railway systems have network effects. A change at one location can result in changes elsewhere. The paper describes how to identify and measure network effects. Network effects are measured by queuing time for the trains and/or passengers.


When railway capacity and delays are investigated, the analysis is often restricted to a single railway line or section of the network. Railway operation however often has network effects as a change in one part of the network can influence other parts of the network. This may even be far away from where the original change was made. These network effects are due to the fact that train routes (often) are quite long and that the railway system has a high degree of interdependencies as trains cannot cross each other or overtake each other everywhere in the network.

Network effects are dependent on the given infrastructure and timetable and can result in longer travel times for trains and passengers. The passengers can be further affected of the network effects due to the fact that not all the wanted correspondences to/from other trains can be kept due to too many interdependencies ? or network effects. Furthermore, the network effects can result in reduced capacity as some trains or train routes can make it impossible to run other trains or train routes. This network effect also affects the passengers as the frequency is reduced.

On rural branch railway lines, where local train lines connect to the national InterCity system, it is unlikely, that any project will influence the national timetable; There is a low degree of network effects. Removing a major bottleneck on a main-line it is on the other hand likely to create changes on many connecting lines, and maybe improve the timetable across the national network; There is a high degree of network effects. The difference in the degree of network effects is due to the train hierarchy where long distance national and international train routes influences more trains (and correspondences) than local train routes and that more trains (normally) are operated on the main lines.

It is possible to evaluate most network effects by measuring the queuing time of the trains (extended running time due to other trains in the network) in the railway system. However, the queuing time of the railway system does only measure the network effects of the trains operated and not the trains which cannot be operated due to lack of suitable train paths. Furthermore, measurement of network effects based on queuing time in the railway system does not describe how the passengers are affected.

Examining the network effects of the passengers using the newly developed passenger punctuality model it is possible to include the trains not operated and the correspondences too. The passenger punctuality model is a route choice model which calculates the difference in travel times for the wanted timetable (for the optimal case) and the actually published timetable where the passengers use the same route of travel. In this way one could say that the queuing time of the passengers are used to evaluate the network effects.


Association for European Transport