Stochastic simulation of the 2023 network-wide timetable

Assessing the performance impact of a major combined timetable and infrastructure change

In Norway the timetable change of 2023 has been in preparation for several years. It combined the entry into service of a new line between Oslo and Ski with a major change in the timetable to take advantage of the new infrastructure which, however, is still being finalised in its connections within Oslo main station. The new line allows to dramatically reduce running times for direct services between Oslo and Ski, and at the same time reduces congestion of the historic line, which has been operating for years significantly over capacity.

Due to the network-wide impacts of these changes, estimating their likely performance required microscopic simulation. TRENOlab was asked to run a network-wide simulation of the R23 timetable to estimate its punctuality compared to the reference R22.

In few weeks trenolab prepared the microscopic model of the network, which had previously been used mostly for deterministic simulations or for testing smaller areas, and configured it to run a network-wide simulation. This task required a comprehensive quality-check in particular of the interlocking system and the configuration of the local dispatchers to handle crossings, delays as well as long freight trains. In Norway, unlike in other countries, freight trains are typically longer than the crossing tracks available for a great number of stations, and thus dispatching them requires a special attention to avoid deadlocks.

The simulation allowed first to identify a few critical spots in the draft timetable, which were addressed by Bane NOR before releasing the final version. Then, it demonstrated the positive impact of R23 on performance on almost all lines, with the remaining ones being unaffected by it.

The quality of results, the time required to obtain them, and the possibility to use the microsimulation approach taking advantage of the experience gained so far, convinced Bane NOR to deploy it for the next timetable changes, setting up an internal process and training their team of treno and trenissimo users on it.

Our tasks:

  • Microscopic simulation model calibration
  • Microscopic stochastic simulation
  • 2022/08 - 2022/11
  • Norway
  • Bane NOR
  • Timetable Planning, Data Analysis
  • TRENOanalysis, TRENOplus, trenissimo