In this thesis the symptoms and underlying behaviour of congestion on railways are analysed and discussed. As well as in many other countries, Sweden faces increasing demand for transportation. To meet this new demand, railways play an important role. Today, the capacity of the Swedish rail network is not upgraded at the pace necessary to keep up with the increase in traffic demand. The sensitivity of the railway system rises as the capacity utilisation increases. At some point maximum capacity is reached when the marginal gain of operating one extra train is lower than the costs in terms of longer travel times and increased sensitivity to delays.

Several different methodologies are employed in this thesis to analyse capacity. The first uses real data from the Swedish rail network, train operation and delays to analyse how different factors influence available capacity and train delays. Several useful key performance indicators are defined to describe capacity influencing properties of the infrastructure and the rail traffic. The rail network is divided into subsections for which the indicators have been estimated. This makes it possible to discern their different characteristics and identify potential weaknesses.

The second approach employs the railway simulation tool RailSys in extensive simulation experiments. This methodology is used to analyse the characteristics of double-track operation. Simulation of several hundred scenarios are conducted to analyse the influence of traffic density, traffic heterogeneity, primary delays and inter-station distance on secondary delays, used timetable allowance and capacity. The analysis gives an in-depth understanding of the mechanisms of railway operation on double-track lines.

A simulation model for strategic capacity evaluation, TigerSim, is developed that can be used to speed up and improve capacity planning and evaluation of future infrastructure and timetables designs on double-track railway lines. For a given infrastructure and plan of operation, the model can be used to generate and simulate a larger number of timetables. This gives two major advantages:

• Using many timetables makes results general
• It is possible to consider both static and dynamic properties of the timetables in the capacity analysis

The first aspect is especially useful in the evaluation of future scenarios as the timetable then often is unknown. The second is an advantage since an improvement in capacity can be measured in a combination of increased frequency of service, shorter travel time and reduced delays. The output of the model can either be used to directly determine capacity from a quality of service perspective, or used as input to cost-benefit analysis (CBA).