At most ~500 m on foot to reach an entrance.
150 km of network ÷ 1 km — exactly the planned budget.
Shallow: simple boring, low ground pressure.
Except at major hubs. Millions saved per year.
The principle: enter from the surface
Rather than a heavy underground box dug in a trench and topped with an elevator, the station is simply a glazed pavilion at street level. It holds the doors, the rental kiosk, the bike racks — and the cycling path that crosses the ground floor then descends into the tunnel down a ramp.
That ramp is not excavated: it is bored, like the main tunnel, from a surface portal. It is the cheapest, simplest thing to build in the whole project — no vertical shaft, no elevator, no closed street. The spiral staircase in the renders is multi-level bike parking, not the through-route: cyclists ride flat and dive into the ramp.
Two example stations
The same principle comes in two configurations depending on the site. Here are both — each with its compatible cross-section, its cost and its advantages.
Example A — Ramps on a deep tunnel (The Boring Company style)
✓ Advantages
- Continuous flow: through-riders never climb
- The network spine stays deep — works near intersections
- Ramps bored in sequence with the machine (cheap junction)
- “Boring Company” aesthetic: tech credibility
To consider
- Two tangential junctions to bore ($2–4M item)
- Separate building + land to plan for
Cost ≈ $10M per station (range $8–13M). The default model on through arteries, mid-segment.
Example B — Building above the tunnel (integrated pavilion / terminus)
✓ Advantages
- Short ramps: the tunnel comes up to the station
- Can fit inside an existing building — saves land and pavilion
- Simple symmetric access, soft and welcoming look
- Perfect at termini and entry points (e.g. “To Downtown Lévis”)
To consider
- The building must sit right above the line (alignment constraint)
- Less suited where the tunnel must stay very deep (near intersections)
Cost ≈ $8M per station ($6–10M), and as low as ~$4–8M if an existing building is reused instead of building a new pavilion.
The geometry: 10 m deep, short ramps
At just 10 m deep, the access ramp is never a major cost item. It all comes down to the chosen grade: the gentler it is, the longer the ramp, but the more accessible to everyone.
| Grade | Length per ramp | User profile |
|---|---|---|
| 5% (≈ 2.9°) | ~200 m | Comfortable for all, even mobility scooters |
| 6% (≈ 3.4°) | ~167 m | The network's sweet spot |
| 8% (≈ 4.6°) | ~125 m | Shorter; e-bikes & cargo bikes at ease |
| 7° (≈ 12.3%) | ~82 m | Minimum length — constrained sites |
The right setting for the network is around 5 to 6% (ramps of ~165 to 200 m): still short, and climbable by the vast majority thanks to the electric assist of the shared bikes. The shallow depth is precisely what allows this gentle grade at no extra cost.
Stations mid-segment, not at intersections
An intersection is where two tunnels cross at different depths (one passes under the other). Grafting a station there would complicate everything. So stations go mid-segment, where the tunnel is straight, alone, at constant depth. The network thus has two clearly separate kinds of point:
- Intersections — deep, continuous, no station, no stop. Pure flow.
- Mid-segments — where the stations live, on a “clean” stretch.
With one station per kilometre and intersections more than 400 m apart, there is always ample room for the two ramps (entry + exit ≈ 330–400 m) without encroaching on the neighbouring crossing.
Why this matters: if the main tunnel itself rose at every station, everyone would climb 10 m at each stop — 50 m of cumulative climb over 5 stations. By keeping the spine deep, only those entering or exiting climb. Station spacing therefore has no effect on the speed of through-riders.
The cost, item by item
At the dossier's rate (CAD $12M/km of tunnel bored in Québec rock), for the full “pavilion + ramps” model at 10 m:
| Item | Estimated cost |
|---|---|
| 2 bored ramps (~165 m total) | ~$2M |
| 2 tangential junctions (ramp ↔ tunnel) | $2–4M |
| Surface building (pavilion) | $1–3M |
| Portals, lining, RFID gates, signage | ~$1M |
| Lighting, ventilation, drainage | ~$1.5M |
| Land (or lease if existing building) | $1–2M |
| Total per station | ~$8 to $13M (≈ $10M) |
The access itself — the ramps — is only ~$2M of that. The bulk of the cost is the building, the junctions and the systems, not the depth. Model B (building above the tunnel, fitted into an existing space) drops even below $8M.
Three models side by side
| Model | Construction | Elevator | Surface works | Experience |
|---|---|---|---|---|
| Underground, side access | ~$6.5–12M | yes | trench (closed street) | go down + wait |
| Underground, central island | ~$6.5–12M | yes | trench | go down + wait |
| Surface + bored ramps | ~$8–13M | no | near zero (boring) | continuous riding |
In raw construction cost, the surface model is not cheaper than an underground station — it is even slightly more, because of the ramp length. The real gain is elsewhere, and it is major.
Why it's the right default model
The elevator is the operational nightmare: constant maintenance, breakdowns, certification. Removing it across ~150 stations means millions saved per year and far higher reliability.
Boring from a portal avoids digging and closing the street (“cut-and-cover”), the hidden cost and political nightmare of urban projects. Fewer delays, permits, objections.
This is exactly how The Boring Company builds its Loop: surface stations, descent down a bored ramp. The project becomes more realistic, not less.
The one real caveat — the grade. A critic will say “you're forcing people to climb 10 m.” The answer: keep the grade at 5–6% (ramps of 167–200 m), practicable for a mobility scooter and trivial for an e-bike — and the network pushes shared e-bikes precisely. For the very few who can neither pedal nor use a scooter, keep one elevator, only at major hubs (Sainte-Foy, U. Laval, Old Québec).
Technical note: the ramps must be bored during the excavation of the main tunnel (a coordinated boring-machine sequence), not afterward in a tunnel already in service — otherwise the junction costs far more.