top of page
Search

Metro 3 in Brussels under uncertainty: scenario-based accessibility analysis

By: Brecht Verbeken, Arne Vanhoyweghen en Vincent Ginis

ree

Research Summary

Metro 3 is planned as a fully automated north–south metro line linking Forest in the south to Evere in the north. For over a decade, the project has been framed as a solution to congestion and urban mobility challenges, yet it has been repeatedly delayed, with projected costs escalating from €0.85 billion in 2009 to nearly €5–7 billion today. Despite the scale of the investment, no publicly available, reproducible analysis has examined what Metro 3 would actually deliver in terms of accessibility: the fundamental ability of residents to reach jobs, services, and opportunities via public transport. Our work aims to close this gap.

We conducted the first scenario-based accessibility appraisal of Metro 3 using General Transit Feed Specification (GTFS) data from all regional public transport operators (STIB/MIVB, De Lijn, TEC, SNCB/NMBS), edited to reflect three possible futures:

  1. Status quo (no metro, tram corridor remains as today).

  2. Partial conversion (southern premetro between Albert and Brussels-North replaced by metro).

  3. Full line (complete metro from Albert to Bordet).


Accessibility was measured between 647 evenly spaced 500 m grid points across the Brussels-Capital Region. For each scenario, we computed door-to-door public transport travel times under nine representative conditions: morning and evening peaks plus weekend midday, each at three departure times (t−10, t, t+10).


Key Findings

  1. System-wide improvements are modest in average terms but significant in distributional terms.

    • Partial conversion yields average travel time reductions of ~16.5 seconds per trip

    • Full build-out doubles the effect: average gains of ~36.1 seconds

  2. The benefits are unevenly distributed and spatially concentrated.

    • Both scenarios strongly benefit Forest and Saint-Gilles, districts with below-average incomes and high reliance on public transport.

    • Only the full line propagates accessibility improvements further north into Schaerbeek and Evere, extending the benefits to a larger set of disadvantaged neighborhoods.

  3. Reliability matters as much as speed.

    • Partial conversion risks degrading stability: small time savings often come with higher variability.

    • Full Metro 3 improves both speed and reliability, reducing risk-adjusted travel times consistently across the city.

  4. Policy interpretation.

    • Time savings alone cannot justify multi-billion euro costs.

    • The case for Metro 3 rests on broader benefits: creating a resilient north–south backbone, reducing socio-spatial inequalities, supporting climate goals, and enabling a coherent redesign of surface transport.


The headline averages of 16.5 or 36.1 seconds per trip can be misleading if read in isolation. Scaled up over more than 400 million annual trips in Brussels, these seemingly small differences accumulate into an enormous volume of saved time. Moreover, our average weights every possible route in the city equally, while in practice the most heavily used flows are concentrated along the north–south corridor, the very axis Metro 3 (partial or full) would serve. T

his means that the realized benefits for actual travelers are likely larger than our uniform statistics suggest. Beyond this, accessibility is a non-ergodic process: a single departure snapshot does not represent the daily experience of an individual traveler, whose journey times vary with schedule fluctuations. By explicitly modeling multiple departure times across weekdays and weekends, we capture this temporal variability and show that the full line consistently improves both speed and reliability, while the partial conversion delivers smaller and less stable gains. A key reason for the partial scenario’s underperformance lies in operational constraints: without a dedicated maintenance and storage depot, rolling stock must be staged from elsewhere each day, limiting capacity and degrading reliability.

Figure 1: Time gain for partial Metro 3 (the greener, the greater the time gain)
Figure 1: Time gain for partial Metro 3 (the greener, the greater the time gain)

Figure 2: Time gain for full Metro 3 (the greener, the greater the time gain)
Figure 2: Time gain for full Metro 3 (the greener, the greater the time gain)

Role of VSC

This research would not have been possible without access to the computational infrastructure of the Flemish Supercomputer Center (VSC). Simulating nine departure conditions across 639 origins and 638 destinations, for three scenarios, with 3 timestamps for each scenario, resulted in millions of routing queries. This project demanded RAM-intensive simulations equivalent to more than four CPU-years of computation.


Read the full publication in BRUZZ media (in Dutch) here

ree

🔍 Your Research Matters — Let’s Share It!

Have you used VSC’s computing power in your research? Did our infrastructure support your simulations, data analysis, or workflow?

We’d love to hear about it!

Take part in our #ShareYourSuccess campaign and show how VSC helped move your research forward. Whether it’s a publication, a project highlight, or a visual from your work, your story can inspire others.

🖥️ Be featured on our website and social media. Show the impact of your work. Help grow our research community

📬 Submit your story: https://www.vscentrum.be/sys

bottom of page