What is being tested?
What happens to traffic when either directional edge of Tunnel Louise has to close for works? The simulation compares normal traffic with two separate cases: closing the tunnel edge toward the city centre, or closing the tunnel edge toward Bois de la Cambre.
To answer this question, we ran a full-day microscopic mobility simulation for Wednesday 19 August 2026, from 00:00 to 23:59. Each scenario represents passenger traffic together with 1,790 scheduled STIB bus movements, allowing the study to examine effects on general traffic and public transport over the morning peak, daytime and evening peak.
The area of interest (AOI) is the footprint of the road network evaluated in the results. It was extracted around both tunnel edges using a configured 0.6 km surrounding buffer. Complete road segments are retained when they cross the extraction boundary, so the resulting footprint extends beyond the nominal buffer. The map below shows this evaluated-network extent; the traffic indicators and impact maps use the same road segments.
How does each directional closure affect traffic?
The two directional tunnel edges are physically separate. Closing one does not mean its traffic can simply be sent through the other: affected drivers must leave the tunnel route and use the surrounding surface streets. Comparing each closure with the same baseline reveals the distinct diversion, congestion and bus impacts that need to be managed for works on that specific edge.
Over the full day, the closure toward Bois de la Cambre has the larger network-wide impact. Compared with normal traffic, average travel time rises by 11.6%, the 95th-percentile travel time by 33.2%, total delay by 60.6%, and average network speed falls by 13.2%. For the closure toward the centre, the corresponding changes are +7.2%, +13.0%, +43.9%, and −9.7%.
When is traffic pressure highest?
Results are reported in 15-minute intervals. The congested-road indicator measures the share of evaluated road length moving below half of its normal free-flow reference speed.
The sharpest centre-bound closure effect occurs from 09:30 to 09:45: 5.7% of evaluated road length is congested, 4.1 percentage points above baseline. The Bois de la Cambre closure peaks from 09:00 to 09:15, at 5.2%, or 3.5 points above baseline.
This distinction matters: the centre-bound closure creates the sharper 15-minute spike, while the Bois de la Cambre closure causes the greater deterioration when results are aggregated over the full day.
Where do conditions deteriorate?
Within the AOI, the maps below show speed changes at each closure’s most disruptive interval. Red roads are slower than in the baseline; blue roads are faster.
During the centre-bound closure peak, the largest slowdowns are concentrated on Avenue Louise and around Rue de Livourne, Rue Washington, Rue du Châtelain, Rue du Bailli and Tunnel Vleurgat. During the Bois de la Cambre closure peak, the strongest effects appear around Avenue Louise, Avenue de la Toison d’Or, Rue des Drapiers, Boulevard de Waterloo, Rue Capitaine Crespel and Place Stéphanie.
Impact on public transport
The public-transport analysis includes bus lines 34, 38, 54, 59, 60, 71, 80, 95 and 96, so the answer concerns both drivers and bus passengers travelling through the wider Louise area.
Across all bus movements, mean running time increases by about 5 seconds with the centre-bound closure and 6 seconds with the Bois de la Cambre closure. The largest line-direction effects are more pronounced: bus 96 toward Gare du Midi is about 70 seconds slower when the centre-bound carriageway closes, while bus 96 toward Legrand is about 71 seconds slower when the Bois de la Cambre carriageway closes. Bus 54 toward Trône also loses roughly 18–20 seconds.
People walking, cycling or using trams are not represented in these traffic results, so their access and safety needs require separate evidence.
Checking the simulation against observed traffic
Twelve traffic-counter locations fall inside the analysis area: three Brussels Mobility counters and nine Telraam cameras. Counter directions with a safe geometric match are linked to the corresponding directed edge in the simulation network; unmatched directions are explicitly flagged. Select a location on the map to compare its 15-minute simulated vehicle flow with observations for Wednesday 19 August 2026, where measurements are available.
Observed vs simulated
Traffic counters
Counters inside the analysis area, projected onto the exact network used by this simulation. Select one to compare measured and simulated flow.
Agreement in the timing and scale of the daily traffic profile increases confidence that the model represents existing conditions at that location. Differences identify where demand or routing needs further calibration. This is a local check at the available counters, not proof that every street in the study area is reproduced accurately; the widget reports missing observations rather than filling the gaps.
What do the results mean for tunnel works?
The simulation does not suggest diverting one tunnel direction through the other. It shows that each directional closure needs its own traffic plan. Closing the edge toward the centre requires particular attention to the sharp 09:30 peak and the Avenue Louise–Bailli–Châtelain area. Closing the edge toward Bois de la Cambre produces the larger full-day increase in travel time and delay, especially around Louise, Toison d’Or, Boulevard de Waterloo and Place Stéphanie.
The figures are comparative simulation results, not exact traffic forecasts. They start from the same demand source and allow 40% of vehicles to reconsider their route during the run. The closure networks ultimately contain 0.7% fewer requested trips toward the centre and 1.3% fewer toward Bois de la Cambre than the baseline, so network-wide totals should be interpreted with that difference in mind. Observed counts, planned work dates and conditions beyond the modelled area should be checked before choosing the closure schedule and the diversion plan for each edge.