Planning Competitive Transportation Networks as High-Speed Rail is Introduced

João Pedro Rosário de Jesus

Key information

Authors:

João Pedro Rosário de Jesus (João Pedro Rosário de Jesus)

Supervisors:

João Pedro Lopes de Albuquerque (João Pedro Lopes de Albuquerque); António Ramos Andrade (António Ramos Andrade)

Published in

July 15, 2025

Abstract

The decarbonization of the transport sector is a growing priority in European mobility policies. Initiatives such as the European Green Deal set ambitious targets to cut emissions, and public debate has intensified around restricting short-haul flights when competitive rail alternatives exist. Congestion problems on major rail corridors and airports highlight the need for more efficient transportation networks. In Portugal, the planned Lisbon-Porto High-Speed Rail (HSR) line is expected to impact travel patterns, particularly on existing air connections. However, limited research has examined how its introduction will affect network dynamics and how competing operators will respond. This research proposes a model to support the design and planning of multimodal transportation networks under such infrastructure transformations. The methodology combines a demand model with an optimization framework formulated as a Mixed-Integer Linear Programming (MILP) problem. The model maximizes the profit of a network involving three operators: an airline, an HSR operator, and a conventional rail operator. It considers itineraries with up to two transfers, allowing combinations across all operators. Environmental targets are addressed through a dedicated constraint that limits total emissions. The model was validated in a simplified illustrative example, confirming its ability to generate and allocate demand to multimodal itineraries while respecting operational and environmental constraints. Its application to the Lisbon-Porto corridor showed that HSR reshaped service configurations, reduced the role of air transport, and shifted demand from less competitive services. These results confirm the model’s ability to support the analysis of infrastructure changes and assist in planning more efficient multimodal networks.